Category · Models & inference
Model APIs and inference for AI agents
Frontier and open-weight model APIs, fast inference hardware and routers that sit in front of many providers. Compared on price, context, tool use, data handling and how often models get retired under you.
Best model APIs and inference for AI agents, the shortlist1090 head-to-head comparisons
Capability keys inference.llm · inference.router · inference.fast · inference.open-weights · All tools
letme.dev/inference.llm picks the top-graded tool in this list and says how to call it direct (how letme works).
The same listing from the live API. Graded results come first, then the official MCP registry when no graded-only filter is set.
https://www.anchorterminal.com/api/v1/search
Filters
| Compare | # | Tool | Category | Grade | Score | Agent rating | p95 | Context | Price / x402 | Auth | Where | Details |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2 | Vertex AI (Gemini Enterprise Agent Platform)Google Cloud · Model API | Models | AA | 91.2 | none | n/a | n/a | from $0.30 / 1M in | OAuth or key | Hosted | ||
|
Google Cloud's enterprise API for Gemini models and Model Garden partner and open models, including Claude, Llama and Qwen, on the platform formerly called Vertex AI, with regional endpoints, provisioned throughput and IAM access. Top strength No incident naming Vertex or Gemini on status.cloud.google.com between 8 July and 6 October 2026 Top weakness The SLA drops from 99.5 to 95 per cent for short-term availability models, Gemini 3.7 and 3.6 Flash among them |
||||||||||||
| 5 | OpenAI APIOpenAI · Model API | Models | AA | 85.8 | 3.5 (8) | n/a | n/a | from $0.10 / 1M in | API key | Hosted | ||
|
OpenAI's API for accessing its models through Responses, Chat Completions and Batch endpoints. Top strength Official OpenAPI document and an llms.txt index Top weakness Elevated errors across the API for about 5 hours 20 minutes on 29 September and about 90 minutes on 17 September 2026 |
||||||||||||
| 6 | GPT-6.2 AstraOpenAI · Model API | Models | AA | 85.8 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
GPT-6.2 Astra is OpenAI's most capable model in its API, for demanding reasoning, coding and computer use, with a 1.05M-token context, 128K output, text and images in, at $8 input and $40 output per million tokens. Top strength $8 input and $40 output per million tokens, 20% below GPT-6 Astra, with cached input at $0.40 Top weakness Tool calling works only through the Responses API, not Chat Completions |
||||||||||||
| 13 | Amazon BedrockAmazon Web Services · Model API | Models | A | 83.5 | none | n/a | n/a | from $0.30 / 1M in | OAuth or key | Hosted | ||
|
AWS's managed API for foundation models from Anthropic, OpenAI, Amazon, Meta, Mistral, DeepSeek, Qwen, xAI and others, called through the Converse API or OpenAI- and Anthropic-compatible endpoints under one AWS account. Top strength One API and one bill for Anthropic, OpenAI, Amazon, Meta, Mistral, DeepSeek, Qwen, xAI and other model families Top weakness An AWS account, IAM setup and a Region choice come before the first call, and an agent can't sign up alone |
||||||||||||
| 22 | Azure OpenAI in Foundry ModelsMicrosoft · Model API | Models | A | 81.2 | none | n/a | n/a | from $0.10 / 1M in | OAuth or key | Hosted | ||
|
Microsoft Azure's hosted API for OpenAI's GPT models and other Foundry Models sold by Azure, such as DeepSeek, Llama, Mistral and Grok, called through an OpenAI-compatible v1 endpoint on each customer's Azure resource. Top strength Public OpenAPI 3.2 spec for the v1 API, 71 paths covering Responses, Chat Completions, batches, files and vector stores Top weakness Intermittent failures and 5xx errors in Sweden Central for about six hours on 29 September 2026 |
||||||||||||
| 38 | Claude APIAnthropic · Model API | Models | A | 79.2 | 3.5 (4) | n/a | n/a | from $0.10 / 1M in | API key | Hosted | ||
|
Anthropic's Messages API for Claude, with server-side tools, an MCP connector and computer use. Top strength Structured outputs and strict tool use are GA, with grammar-constrained sampling on every current model Top weakness Many API incidents labelled major in the 90 days to 1 November 2026, among them a critical one on 23 October and five more from 24 to 27 October |
||||||||||||
| 39 | Claude Fable 5.2Anthropic · Model API | Models | A | 79.2 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Claude Fable 5.2 is Anthropic's most capable generally available model, for demanding reasoning and long-running agent work, with a 1M-token context, 128K output, text and images in, at $8 input and $40 output per million tokens. Top strength $8 input and $40 output per million tokens, 20% below Fable 5.1, with cache reads at 2.5% of input Top weakness Adaptive thinking is always on and can't be turned off |
||||||||||||
| 42 | BlockRun.AIBlockRun, Inc. · MCP server | Models | A | 78.9 | 3.5 (2) | n/a | n/a | Pay per use x402 | OAuth or key | Local | ||
|
Model and API gateway with an OpenAI-compatible interface, MCP support and per-request stablecoin payments through x402. Top strength x402 v2 on its own gateway, USDC on Base, Solana and Arc, with no account needed Top weakness Since 21 October 2026 prompts and completions sent through the Base, Arc and API gateways are stored with no fixed expiry |
||||||||||||
| 71 | Claude Mythos 5.2Anthropic · Model API | Models | BB | 77.3 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Claude Mythos 5.2 is the same model as Claude Fable 5.2 without Fable's stronger cyber and biology safeguards, for organisations Anthropic verifies, with a 1M-token context, text and images in, at $8 and $40 per million tokens. Top strength $8 input and $40 output per million tokens, 20% below Fable 5.1, with cache reads at 2.5% of input Top weakness Only for organisations Anthropic verifies, so an agent can't start on it alone |
||||||||||||
| 80 | GroqCloudGroq · Model API | Models | BB | 77 | 3.5 (8) | n/a | n/a | from $0.075 / 1M in | API key | Hosted | ||
|
OpenAI-compatible inference API serving open-weight models on Groq's processors. Top strength Free plan with no card, at 30 requests a minute and 1,000 a day on gpt-oss Top weakness Four model shutdown dates between 2026-07-17 and 2026-09-21, with no stated minimum notice |
||||||||||||
| 202 | xAI API (Grok)SpaceXSI (xAI) · Model API | Models | BB | 72.9 | none | n/a | n/a | from $1.25 / 1M in | API key | Hosted | ||
|
xAI's API for its Grok models (the docs now name the company SpaceXSI), with text, image, video and voice endpoints, server-side search and remote MCP tools, through an OpenAI-compatible REST API or gRPC. Top strength Grok 4.7 at $2/$6 per million tokens below 200K prompt tokens, with a 500K context and 20% off through the Batch API Top weakness API models outage of about 3 hours 40 minutes on 3 September 2026, and high error rates on grok-4.6 and grok-4.5 on 17 August |
||||||||||||
| 246 | Fireworks AI InferenceFireworks AI · HTTP API | Models | BB | 71.8 | 4.3 (4) | n/a | n/a | Freemium | API key | Hosted | ||
|
Fireworks AI Inference runs open models through OpenAI-compatible chat and completion APIs. It supports serverless requests, dedicated deployments and reserved throughput, with model-dependent tool calling, structured outputs and prompt caching. Top strength An inference-user role limits service accounts to calling models, with no rights to create or change resources Top weakness Cloud provider outages on 11 August and 14 October hit many serverless models for hours |
||||||||||||
| 288 | Cloudflare Workers AICloudflare · Model API | Models | BB | 70.6 | 3.0 (4) | n/a | n/a | from $0.0605 / 1M in | API key | Hosted | ||
|
Workers AI is Cloudflare's serverless inference service for open-weight models, covering text generation, embeddings, images and speech. Agents call it through the Cloudflare REST API, OpenAI-compatible endpoints or a Workers binding. Top strength Per-model token prices, cached-input prices and neuron rates are published without a login, with 10,000 neurons a day free on the Workers Free plan Top weakness No SLA for Workers AI was found in the docs or the agreements read |
||||||||||||
| 306 | Cerebras InferenceCerebras Systems Inc. · HTTP API | Models | B | 69.9 | 3.3 (4) | n/a | n/a | Paid | API key | Hosted | ||
|
Cerebras Inference serves language models on Cerebras wafer-scale processors through an OpenAI-compatible API. It supports streaming, structured outputs, tool calling and prompt caching, with shared and dedicated inference options. Top strength Public OpenAPI 3.1, Markdown docs, and official Python and TypeScript SDKs Top weakness Trial access requires a verified payment method and ends after $5 or 30 days |
||||||||||||
| 344 | Cloudflare AI GatewayCloudflare · Model router | Models | B | 69.2 | 3.0 (4) | n/a | n/a | 5% fee | API key | Hosted | ||
|
AI Gateway is Cloudflare's proxy and router for model providers. Agents call OpenAI-style, Anthropic-style or native endpoints on the Cloudflare API, with logging, caching, rate and spend limits, retries, fallbacks and one prepaid credit balance. Top strength Core gateway functions (analytics, caching, rate limiting) are free on all plans, and provider prices pass through with no markup Top weakness API tokens are account-scoped. Any token with AI Gateway Run can send requests through every gateway in the account and spend its stored provider keys |
||||||||||||
| 347 | Together AI InferenceTogether AI · HTTP API | Models | B | 69.1 | 3.0 (4) | n/a | n/a | Paid | API key | Hosted | ||
|
Together AI Inference serves open models through an OpenAI-compatible API, with serverless requests, batch processing, dedicated model deployments and provisioned throughput. Model-dependent capabilities include tool calling, structured outputs and vision. Top strength Public OpenAPI 3.1, Markdown documentation and Python and TypeScript SDKs Top weakness A $15 credit purchase is required before platform access; no free trial |
||||||||||||
| 349 | OpenRouterOpenRouter · Model router | Models | B | 69.1 | 3.0 (2) | n/a | n/a | 5.5% fee | API key | Hosted | ||
|
One OpenAI-compatible API in front of 80+ model providers, at the providers' own token prices, with fallbacks and routing controls. Top strength Public OpenAPI document, llms.txt with Markdown twins, and a changelog generated from the spec with migration notes Top weakness No machine payment. USDC top-ups go to a prepaid balance |
||||||||||||
| 352 | Vercel AI GatewayVercel · Model router | Models | B | 69 | 3.0 (4) | n/a | n/a | $30 / mo | OAuth or key | Hosted | ||
|
Vercel AI Gateway is a hosted router for model APIs. One API key or Vercel OIDC token reaches models from OpenAI, Anthropic, Google and others through OpenAI-compatible, Anthropic-compatible and AI SDK interfaces, with budgets, request logs and provider failover. Top strength Token prices are each provider's list price with no markup or platform fee, including requests made with the customer's own provider keys Top weakness The free monthly credit needs a valid payment method on the team, and its amount is not stated in the reviewed docs |
||||||||||||
| 386 | Mistral AI APIMistral AI · Model API | Models | B | 68.2 | 4.0 (2) | n/a | n/a | from $0.10 / 1M in | API key | Hosted | ||
|
Mistral's API for its open-weight and proprietary models, with EU and US regional endpoints. Top strength Public OpenAPI 3.1.0 document at docs.mistral.ai/openapi.yaml and an llms.txt with Markdown twins Top weakness Labs and preview models, Large 4 among them, may train on data sent from 2026-09-25, whatever the opt-out or zero-retention setting |
||||||||||||
| 404 | Cohere Chat API (Command models)Cohere · Model API | Models | B | 67.9 | 3.0 (4) | n/a | n/a | from $0.0375 / 1M in | API key | Hosted | ||
|
Cohere's Chat API generates text with the Command family of models over REST (v2 chat), with tool use, structured outputs, citations and streaming. SDKs cover Python, TypeScript, Java and Go. Top strength Public OpenAPI 3.1 file in the docs repository, last changed 29 September 2026, and a Markdown twin of every docs page Top weakness Production keys work like trial keys on Command A+, Reasoning, Translate and Vision. The docs send production use of those models to sales or to Model Vault |
||||||||||||
| 432 | SambaCloudSambaNova Systems, Inc. · Model API | Models | B | 67.4 | 3.5 (4) | n/a | n/a | from $0.22 / 1M in | API key | Hosted | ||
|
SambaCloud is SambaNova's hosted inference API for open-weight models running on its own RDU processors. It answers OpenAI-style chat, completions and Responses calls and Anthropic-style Messages calls, with Python and TypeScript SDKs. Top strength Public OpenAPI 3.1.1 document for all 10 operations, plus Top weakness Production models get a notice of two to three weeks, and 13 model ids were removed between 9 March and 9 June 2026 |
||||||||||||
| 484 | LiteLLMBerriAI · Model router | Models | B | 66.2 | none | n/a | n/a | Freemium | API key | Local | ||
|
Open-source Python SDK and self-hosted AI gateway from BerriAI that puts 100+ model providers behind one OpenAI-compatible API, with virtual keys, spend tracking, budgets, fallbacks, guardrails and an MCP gateway. Top strength MIT core, self-hosted, with no telemetry when self-hosted per the data security page Top weakness litellm 1.82.7 and 1.82.8 on PyPI carried a credential stealer for about 40 minutes on 24 March 2026 |
||||||||||||
| 538 | DeepInfraDeep Infra Inc. · Model API | Models | B | 65 | 3.5 (4) | n/a | n/a | from $0.06 / 1M in | API key | Hosted | ||
|
DeepInfra is a hosted inference API for open-weight and some third-party models, covering chat, embeddings, reranking, image, video and speech. It answers OpenAI-style and Anthropic-style calls at api.deepinfra.com with a Bearer key. Top strength
Top weakness A deprecated model gets at least one week's notice, and requests are then forwarded to a replacement model under the old id. Eight models were redirected or scheduled between 8 October and 3 November 2026 |
||||||||||||
| 547 | Gemini 4 HydrogenGoogle · Model API | Models | B | 64.9 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Gemini 4 Hydrogen is Google's fast, lower-cost Gemini 4 model for everyday coding and agent work, on every tier of the Gemini API with a 1,048,576-token limit shared between input and output. Top strength On the free tier of the Gemini API, unlike Carbon and Argon Top weakness temperature, top_p and top_k are ignored, and thinking_budget, MINIMAL, penalties and extra candidates return a 400 |
||||||||||||
| 559 | Gemini Developer APIGoogle · Model API | Models | B | 64.7 | 3.5 (2) | n/a | n/a | from $0.25 / 1M in | API key | Hosted | ||
|
Google's API for Gemini models, including content generation and agent interactions. Top strength Free tier on 3.8 Flash and Flash-Lite with no billing account Top weakness Free-tier prompts and responses improve Google products and may be read by human reviewers |
||||||||||||
| 596 | Tencent Hy4Tencent (Hy team) · Model API | Models | B | 64 | none | n/a | n/a | Pay per use | API key | Local | ||
|
Tencent Hy4 is Tencent's flagship language model, released on 4 November 2026 with image understanding and computer use, served on Tencent Cloud TokenHub as hy4 with a 1M-token context, at ¥6 input and ¥18 output per million tokens. Top strength 1M-token context with 960K input and 64K output, per TokenHub's model list Top weakness Weights, licence and system card are promised but not published; tencent/Hy4 didn't answer on Hugging Face on 11 November |
||||||||||||
| 600 | Gemini 4 CarbonGoogle · Model API | Models | B | 63.9 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Gemini 4 Carbon is Google's strongest Gemini model, for long agentic coding, cybersecurity and research work, on the paid tier of the Gemini API with a 1,048,576-token limit shared between input and output. Top strength Introductory price of $2 input and $10 output per million tokens, with cached input 95% off Top weakness Paid tier only, with no free tier to try it on |
||||||||||||
| 696 | Eden AIEden AI SAS · Model router | Models | C | 61.8 | none | n/a | n/a | 5.5% fee | API key | Hosted | ||
|
Eden AI is a French AI gateway with one OpenAI-compatible API key for about 1,000 chat and expert models from 80 or more providers, including OCR, speech, translation and image analysis, with an EU-only endpoint and a hosted MCP server. Top strength OpenAI-compatible v3 API with an OpenAPI 3.1 spec and a keyless model catalogue with prices and regions Top weakness No public status page, incident history or SLA found |
||||||||||||
| 701 | Tencent Cloud TokenHubTencent Cloud · Model router | Models | C | 61.7 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Tencent Cloud TokenHub is a hosted model gateway serving Tencent's Hy models and third-party models such as DeepSeek, Kimi, GLM and MiniMax through OpenAI-compatible and Anthropic-compatible endpoints in Guangzhou and Singapore, with per-key scopes and quotas. Top strength API keys can be limited to chosen models and inference services, up to 200 source IP ranges, and daily, monthly or total token quotas Top weakness Sign-up needs a Tencent Cloud account with real-name verification. The international site answered with a bot check, so access from outside China was not established |
||||||||||||
| 706 | Nemotron 3.5 Super VLNVIDIA · Model API | Models | C | 61.6 | 3.0 (4) | n/a | n/a | from $0.15 / 1M in | API key | Hosted | ||
|
Nemotron 3.5 Super VL is NVIDIA's open-weight vision-language model for agents, a 120B mixture of experts with 12B active parameters that reads text, images and video. NVIDIA hosts a free trial endpoint, and the weights are under OpenMDW-1.1. Top strength OpenAPI 3.1 file for the hosted chat endpoint, with enums for Top weakness NVIDIA says its hosted endpoints are for prototyping, research and testing only, and production use needs a partner endpoint or self-hosting |
||||||||||||
| 741 | Prism InferencePrism Technologies Inc · Model API | Models | C | 60.8 | 3.5 (4) | n/a | n/a | from $0.09 / 1M in | API key | Hosted | ||
|
Prism is a hosted inference API from Prism Technologies Inc for open-weight models, aimed at coding agents. It accepts OpenAI Chat Completions, OpenAI Responses and Anthropic Messages requests at api.prisminference.com. It launched on 24 September 2026. Top strength One key works across OpenAI Chat Completions, OpenAI Responses and Anthropic Messages, with a public OpenAPI 3.1 file, llms.txt and Markdown docs Top weakness Two models. The docs mark Gemma 4 31B as request access per organisation, while llms.txt lists it as available |
||||||||||||
| 891 | AntseedAntseed Foundation · Model router | Models | C | 57 | none | n/a | n/a | 4% fee | None | Local | ||
|
Antseed is an open-source peer-to-peer network for buying and selling AI inference. A buyer runs a local proxy with OpenAI and Anthropic format endpoints, which routes each request to an independent seller and pays in USDC on Base. Top strength No sign-up. Top weakness The buyer proxy copied client credential headers to sellers until v0.2.58 (12 October 2026), with no published advisory |
||||||||||||
| 917 | MiniMax APIMiniMax · Model API | Models | C | 56.5 | 2.8 (4) | n/a | n/a | from $0.15 / 1M in | API key | Hosted | ||
|
MiniMax's API for its M-series text models, M3.1 Flash and M3, in the Anthropic Messages, OpenAI Chat Completions and OpenAI Responses formats, with 1M-token context and automatic prompt caching. Top strength M3.1 Flash at $0.15 in, $0.60 out and $0.015 for cache reads per million tokens, flat at any input length Top weakness The terms let MiniMax use inputs and outputs to develop and improve its services, and no opt-out was found |
||||||||||||
| 922 | Muse Spark OpenMeta · Model API | Models | C | 56.5 | none | n/a | n/a | from $1.25 / 1M in | None | Local | ||
|
Muse Spark Open is Meta's open-weight multimodal model for coding and agentic work, with 816B total and 42B active parameters and a 1M-token context, released on 15 October 2026 under the Muse Model Licence for self-hosting. Top strength 816B total and 42B active parameters with a 1,048,576-token context, taking text, image, video and audio input Top weakness The NVFP4 checkpoint (about 537 GB) serves on four B300 GPUs and the BF16 one (about 1.65 TB) on eight |
||||||||||||
| 938 | Router.com (Ramp Router)Ramp · Model router | Models | C | 56.2 | none | n/a | n/a | 0% fee | API key | Hosted | ||
|
Router.com (Ramp Router) is Ramp's LLM gateway. One API key reaches OpenAI, Anthropic, xAI and open-weight models through OpenAI Responses, Chat Completions and Anthropic Messages endpoints, with fallbacks, Flex-tier routing, shadow testing and benchmark aliases. Top strength llms.txt and a Markdown copy of every docs page, with a field table for each endpoint and an error reference with an action per code Top weakness Ramp-side content recording is on by default and keeps prompts, outputs and tool calls for one year |
||||||||||||
| 940 | Novita AINovita AI · Model API | Models | C | 56.1 | none | n/a | n/a | from $0.05 / 1M in | API key | Hosted | ||
|
Novita AI is a hosted inference API for open-weight language models, with embeddings, reranking, image, video and speech models on the same key. It answers OpenAI-style calls at api.novita.ai with a Bearer key. Top strength API keys take an expiry of 24 hours, 30 days or 90 days, a per-key model access policy and a per-key source IP allowlist, the last two settable by API Top weakness No status page or SLA was found on the home page, the docs index or the docs site map. The home page states 99.5% uptime with no supporting document |
||||||||||||
| 947 | Jais 3Institute of Foundation Models (MBZUAI) and Inception42 · Model API | Models | C | 55.9 | 2.5 (4) | n/a | n/a | Free | None | Local | ||
|
Jais 3 is an open-weight family of Arabic and English reasoning models from MBZUAI's Institute of Foundation Models and Inception42, from 0.9B to 375B parameters with up to 512K tokens of context, released under Apache-2.0 for self-hosting. Top strength Six checkpoints from 0.9B (2.2 GB) to 375B-A23B (758 GB), with FP8, GGUF and MLX builds from IFM Top weakness Every documented vLLM, SGLang, Transformers and MLX command passes |
||||||||||||
| 966 | Hugging Face Inference ProvidersHugging Face · Model router | Models | C | 55.3 | none | n/a | n/a | $2 / mo | API key | Hosted | ||
|
Hugging Face's serverless router that sends model calls to partner inference providers such as Groq, Cerebras, Together and Fireworks through one Hugging Face token, with an OpenAI-compatible API at router.huggingface.co/v1 and Python and JavaScript clients. Top strength router.huggingface.co/v1/models lists every routed model with each provider's price, context, tool support, latency and throughput, without a token Top weakness No rate limits, 429 behaviour or error responses documented for the router |
||||||||||||
| 1003 | Beam 1.0Reflection · Model API | Models | C | 54.1 | 2.5 (4) | n/a | n/a | Freemium | API key | Hosted | ||
|
Beam 1.0 is Reflection's open-weight text model for coding, reasoning and agentic work, with 501B total and 23B active parameters and a 1M-token context, served by the Reflection API and released as Apache-2.0 weights. Top strength OpenAI-compatible Chat Completions with Top weakness No SLA, and per-organisation rate limits and the daily token allowance are not published as numbers |
||||||||||||
| 1021 | Meta Model APIMeta · Model API | Models | D | 53.6 | none | n/a | n/a | from $1.25 / 1M in | API key | Hosted | ||
|
Meta Model API is Meta's pay-as-you-go API for its hosted Muse models, including Muse Wonder and Muse Spark, through OpenAI-style Responses and Chat Completions endpoints and an Anthropic-style Messages endpoint, plus image, speech and segmentation routes. Top strength Muse Wonder and Muse Spark take text, image, video, audio and PDF input with a 1,048,576-token context window Top weakness Muse Wonder is labelled a preview, and no deprecation policy or notice period was found |
||||||||||||
| 1038 | Muse EmberMeta · Model API | Models | D | 53.1 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Muse Ember is Meta's smallest and most efficient hosted model, built for sub-agent tasks, tool calls, routing and work over large collections of files and images, taking text, images, video, audio and PDF with a 1M-token context. Top strength $0.10 input, $0.004 cached and $0.50 output per million tokens up to 200K input, published without a login Top weakness No weights and no parameter count published, so it can't be run locally |
||||||||||||
| 1053 | RouterPlusRouterPlus · Model router | Models | D | 52.8 | none | n/a | n/a | 0% fee | API key | Hosted | ||
|
RouterPlus is an LLM gateway that puts Claude, GPT, open-weight, image, video and decision models behind one API key and a prepaid balance, with OpenAI-compatible and Anthropic-compatible endpoints. Top strength llms.txt, llms-full.txt and a Markdown copy of every page, with an integration guide written for coding agents Top weakness No terms of service, privacy policy, status page, changelog or security.txt found for the gateway |
||||||||||||
| 1194 | DeepSeek APIDeepSeek · Model API | Models | D | 47.9 | 2.0 (2) | n/a | n/a | from $0.30 / 1M in | API key | Hosted | ||
|
Hangzhou lab's API for the V4 model family, in the OpenAI Chat Completions, OpenAI Responses and Anthropic request formats, with off-peak pricing. Top strength Accepts the OpenAI Chat Completions, OpenAI Responses and Anthropic request formats Top weakness 36 status incidents from 22 July to 15 October 2026, with partial API outages of three hours on 14 September and 4 hours 52 minutes on 15 October |
||||||||||||
| 1249 | StepFun APIStepFun · Model API | Models | E | 45.6 | 3.3 (4) | n/a | n/a | from $0.10 / 1M in | API key | Hosted | ||
|
The StepFun API is StepFun's paid model API for its Step language, vision and speech models, including the flagship Step 5 Preview, through OpenAI-style Chat Completions and Responses endpoints and an Anthropic-style Messages endpoint. Top strength Step 5 Preview takes text, images and video with a 1M-token context and up to 64k output tokens Top weakness Step 5 Preview is labelled a preview, and no release date or support commitment was found |
||||||||||||
| 1257 | SiliconFlowSiliconFlow Labs Pte. Ltd. · Model API | Models | E | 45.2 | 2.5 (4) | n/a | n/a | from $0.15 / 1M in | API key | Hosted | ||
|
SiliconFlow is a hosted inference API for open-weight models, covering chat, vision, embeddings, reranking, image, video and speech. It answers OpenAI-style and Anthropic-style calls at api.siliconflow.com with a Bearer key. Top strength Per-token, per-image and per-video prices are public for every listed model, such as DeepSeek-V4.1-Flash at $0.15 in and $0.60 out per 1M tokens Top weakness No status page, incident history or SLA was found, and the terms disclaim any uptime or availability commitment |
||||||||||||
| 1387 | SWE-2 FlashCognition AI, Inc. · Model API | Models | F | 37.6 | none | n/a | n/a | Freemium | OAuth or key | Local | ||
|
SWE-2 Flash is Cognition's small, fast coding model, mid-trained and reinforcement-trained from Thinking Machines' Inkling-Small (276B total, 12B active parameters). It runs inside Devin Desktop and Devin CLI, with no model API of its own and no public weights. Top strength Free in Devin Desktop and Devin CLI through 30 November 2026, and selectable per run with Top weakness No model API, OpenAPI entry or model card. The Devin API's session agents list SWE-2 but not SWE-2 Flash |
||||||||||||
| 1394 | GlimpseElorian · Model API | Models | F | 36.7 | 2.5 (4) | n/a | n/a | from $0.60 / 1M in | API key | Hosted | ||
|
Glimpse is Elorian's visual reasoning model, released as a research preview on 16 October 2026. It takes text, images and video and returns text, through Elorian's own API and chat app and through OpenRouter. Top strength Takes text, images and video and returns text, per Elorian's launch post and Together's model page Top weakness Labelled a research preview, with no public API docs, model card, changelog or deprecation policy from Elorian |
||||||||||||
| 1398 | SimplismartVerute Technologies Private Limited · HTTP API | Models | F | 36.4 | none | n/a | n/a | Pay per use | API key | Hosted | ||
|
Simplismart is an inference platform for open models. It runs 150+ models on shared pay-per-token endpoints with an OpenAI-compatible API, and handles dedicated GPU deployments, bring-your-own-cloud clusters and fine-tuning jobs. Top strength Public per-unit prices without a login, such as Llama 3.1 8B at $0.13 per 1M tokens and Whisper v3 Turbo at $0.0018 per audio minute Top weakness No request or token rate limits, 429 handling or retry guidance found in the docs; only GPU quotas, given as 1 and as 8 H100s on two pages |
||||||||||||
Nothing matches these filters. .
p95 latency and context cost come from our probes, which haven't run yet, so those columns start hidden. Grades run from AA to F, and agent-ready means BB or better. Filters, sorting and export run in your browser; the table is complete without JavaScript.
Indexed, not reviewed (295)
Listings sorted into this category from public catalogues (the official MCP registry, APIs.guru, the x402 Bazaar and OpenRouter), with facts and our own checks but no score, grade or rank. How the index works.
| Listing | Kind | What it does | Why it's here |
|---|---|---|---|
| 3DPACK.ING — Container & Truck Load Planning 3dpack.ing | MCP server | Container load planning in Claude & ChatGPT: what fits, how full, a labelled 3D plan in the chat. | vendor's own |
| 3dstreet 3dstreet.app | MCP server | Drive an open 3DStreet scene tab from Claude Desktop or Claude Code via MCP tool calls. | vendor's own |
| AdButler adbutler.com | MCP server | Manage AdButler campaigns, zones, ad items, VAST, and reports from Claude, ChatGPT, and Cursor. | vendor's own |
| adtest-mcp adtest.ai | MCP server | AI-only ad scoring for Claude: score image, video or text ads on 13 dimensions before you spend. | vendor's own |
| AgentDrive agentdrive.sh | MCP server | Capture coding-agent sessions and resume them in any other agent: Claude Code, Codex, Cursor, more. | vendor's own |
| Aion AionLabs | model family | Aion 3.5 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It uses… | on OpenRouter |
| Aion Mini AionLabs | model family | Aion 3.5 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It… | on OpenRouter |
| Aion-2 AionLabs | model family | Aion-2.0 is a variant of DeepSeek V3.2 optimized for immersive roleplaying and storytelling. It is particularly strong… | on OpenRouter |
| Aion-3 AionLabs | model family | Aion-3.0 is a multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. It uses… | on OpenRouter |
| Aion-3.0-Mini AionLabs | model family | Aion-3.0 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the DeepSeek family of… | on OpenRouter |
| Aion-RP AionLabs | model family | Aion-RP-Llama-3.1-8B ranks the highest in the character evaluation portion of the RPBench-Auto benchmark, a… | on OpenRouter |
| ArtifactSend artifactsend.com | MCP server | Share docs, decks, HTML and Claude artifacts as tracked links and data rooms. See who read them. | vendor's own |
| augments.dev MCP server augments.dev | MCP server | Augments MCP Server - A comprehensive framework documentation provider for Claude Code | vendor's own |
| Auten auten.ai | MCP server | Computer use for Claude Code, Codex, Cursor or any MCP client: your computer + Android phone (beta) | vendor's own |
| AXF smartergpt.dev | MCP server | Agent eXoskeleton Framework control plane for workspace-native agent capabilities over MCP. | vendor's own |
| Axis useaxis.dev | MCP server | Coding agents from Claude Code, Cursor and Codex claim jobs and lock files on one shared board. | vendor's own |
| baton kasabeh.tech | MCP server | Pass the baton between coding agents: convert sessions across Claude Code, OpenCode, Codex & more. | vendor's own |
| Beam Reflection | model family | Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows. It uses a sparse… | on OpenRouter |
| ChatGPT Ads getmcpads.com | MCP server | ChatGPT Ads MCP server: 30 read tools, 28 opt-in preview-first writes. | vendor's own |
| Claude Fable Anthropic | model family | Claude Fable 5.2 is Anthropic's most capable model for the most demanding technical work, with gains over Fable 5.1 in… | on OpenRouter |
| Claude FAF faf.one | MCP server | Persistent project context for Claude. IANA-registered .faf format. | vendor's own |
| Claude Haiku Anthropic | model family | Claude Haiku 5.5 is Anthropic's small, fast model for high-volume, cost-sensitive work such as summarization,… | on OpenRouter |
| Claude Opus Anthropic | model family | Claude Opus 5.5 is Anthropic's flagship model for demanding reasoning, coding, and long-horizon agentic work,… | on OpenRouter |
| Claude Sonnet Anthropic | model family | Claude Sonnet 5.5 is Anthropic's Sonnet-class model for well-scoped everyday work, succeeding Claude Sonnet 5 as a… | on OpenRouter |
| Cleanor cleanor.app | MCP server | Zero-auth MCP: image optimize, cited storage/format data, and dev utilities LLMs get wrong. | vendor's own |
| Code Pathfinder codepathfinder.dev | MCP server | Code intelligence MCP server: call graphs, type inference, and symbol search for Python/Go. | vendor's own |
| Codestral Mistral | model family | Mistral's cutting-edge language model for coding released end of July 2025. Codestral specializes in low-latency,… | on OpenRouter |
| Command A Cohere | model family | Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance… | on OpenRouter |
| Command A+ Cohere | model family | Command A+ is Cohere's flagship model for enterprise agentic workflows. It accepts text and image inputs with a 192K… | on OpenRouter |
| Command R Cohere | model family | command-r-08-2024 is an update of the [Command R](/models/cohere/command-r) with improved performance for multilingual… | on OpenRouter |
| Command R+ Cohere | model family | command-r-plus-08-2024 is an update of the [Command R+](/models/cohere/command-r-plus) with roughly 50% higher… | on OpenRouter |
| Command R7B Cohere | model family | Command R7B (12-2024) is a small, fast update of the Command R+ model, delivered in December 2024. It excels at RAG,… | on OpenRouter |
| crisphive.com MCP server crisphive.com | MCP server | Field operations on a deterministic solver — run jobs, crews & fleet from Claude or ChatGPT. | vendor's own |
| Cydonia V4 TheDrummer | model family | Uncensored and creative writing model based on Mistral Small 3.2 24B with good recall, prompt adherence, and… | on OpenRouter |
| DeepSeek V3 DeepSeek | model family | DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and… | on OpenRouter |
| DeepSeek V3 Terminus DeepSeek | model family | DeepSeek-V3.1 Terminus is an update to [DeepSeek V3.1](/deepseek/deepseek-chat-v3.1) that maintains the model's… | on OpenRouter |
| DeepSeek V4 Flash DeepSeek | model family | DeepSeek V4.1 Flash 1017 is the October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the… | on OpenRouter |
| DeepSeek V4 Flash Vision DeepSeek | model family | DeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of [DeepSeek V4 Flash… | on OpenRouter |
| DeepSeek V4 Pro DeepSeek | model family | DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro. | on OpenRouter |
| Devstral Mistral | model family | Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter… | on OpenRouter |
| DidYouSEO didyouseo.com | MCP server | SEO audits, Search Console, GA4, keyword and backlink data for Claude, ChatGPT and Cursor. | vendor's own |
| digline digline.dev | MCP server | Read, measure and explain LLM eval runs of a digline suite. Cannot promote a baseline. | vendor's own |
| Dropl MCP dropl.io | MCP server | Create client-editable photo and video galleries from Cursor or Claude Code, then embed them. | vendor's own |
| Elmo elmohq.com | MCP server | Track prompts on ChatGPT/Claude and all other models to measure AI visibility and perform AEO. | vendor's own, widely used |
| Ember-1 Fireworks | model family | Ember-1 is a specialized reasoning model from Fireworks Research, built on [Kimi… | on OpenRouter |
| ERNIE-Light-DT-1 Baidu | model family | ERNIE-Light-DT-1.0 is based on open-source GLM models and includes research findings from Baidu Search's Lightning… | on OpenRouter |
| etincel-nonfiction etincel.ai | MCP server | Trainable non-fiction writing voice, presets, and an anti-AI-tells audit for Claude, via MCP. | vendor's own |
| Extract ScaleDown | model family | Extract is a description-driven entity extraction model from ScaleDown. Instead of a fixed set of entity types, callers… | on OpenRouter |
| FAQai faqai.app | MCP server | Turn any document into a production-ready RAG dataset from Claude, Cursor, or any MCP host. | vendor's own |
| flowcontent.io MCP server flowcontent.io | MCP server | Connect Claude, Cursor, or any MCP client to your FlowContent account | vendor's own |
| Fugu Max Sakana | model family | Fugu Max is the cost-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a… | on OpenRouter |
| Fugu Ultra Sakana | model family | Fugu Ultra is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu is a… | on OpenRouter |
| Fugu Ultra v2 Sakana | model family | Fugu Ultra v2 is the higher-performance model in Sakana AI's Fugu family. Rather than a single monolithic model, Fugu… | on OpenRouter |
| gate spideriq.ai | MCP server | SpiderIQ Gate: SpiderGate LLM gateway (completions, models, usage, traces) | vendor's own |
| Gemini Argon | model family | Gemini 4 Argon is Google's flagship reasoning model, built for complex software engineering, deep mathematical and… | on OpenRouter |
| Gemini Carbon | model family | Gemini 4 Carbon is a frontier Gemini 4 model from Google for coding, agentic and complex reasoning work. It takes text,… | on OpenRouter |
| Gemini FAF faf.one | MCP server | Persistent project context for Google Gemini. Python/FastMCP. IANA-registered .faf format. | vendor's own |
| Gemini Flash | model family | Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software… | on OpenRouter |
| Gemini Flash Lite | model family | Gemini 3.5 Flash Lite is a high-efficiency model from Google with upgraded agentic capabilities. It is suited for… | on OpenRouter |
| Gemini Hydrogen | model family | Gemini 4 Hydrogen is the high-efficiency workhorse model of the Gemini 4 family from Google, built for high-throughput… | on OpenRouter |
The first 60 of 295, most used first. All 295 are in this page's JSON twin and, with their facts, in /api/v1/indexed.json (category).
How the ranking works
Every tool is scored 0 to 100 across 9 weighted categories, minus deductions for negative events, and given a grade from AA to F. Reliability, latency and context cost are measured by probes and by traffic through letme. Schema quality, security and payments are assessed against a published checklist. Task success comes from a fixed task suite. Payments, including x402, are one category of the nine. Read the methodology.
For companies
Do agents find, use and choose your tools?
An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.