Iris-3B vs SurgePix API

Iris-3B 60

Free Apache-2.0 software with nothing to buy, so 20 + 20 + 20 for pricing, free use and no sign-up. No payment protocol (0). There is no hosted API. Hardware is the owner's cost.

SurgePix API 20

No x402, MPP or L402 (0). Plan prices and monthly credits are public, but credits are spent per token with no per-call price for any API task, so plan-only pricing (10). A free plan with 25 credits a day. The pages don't say whether a card is needed or whether a free account can create API keys, so half, a judgement call (10). A person signs up in a browser and creates the key in settings (0). 0+10+10+0 = 20.

Iris-3B 45

Read as local software, since Iris-3B is a model the owner runs. No package on PyPI. Install is a git clone and pip install -e ., with Python 3.11 or newer, PyTorch 2.7.1 or newer and an NVIDIA GPU with CUDA stated (12 of 20). The repository has 23 pytest files covering the model, flow, checkpoints and parallel training, but no CI workflow, so nothing shows them passing on main (5 of 25). Three GitHub issues in ten days, none a crash, two closed and one open announcing a community ComfyUI port. Six Hugging Face discussions, none a bug report (20 of 25). No tags, releases or changelog. Package version 0.1.0, and new checkpoints are added in folders of one Hub repository with no version pins (3 of 15). Released as a final checkpoint with no beta label, but the package is 0.1.0 and the project ten days old (5 of 15).

SurgePix API 15

Read with the hosted lines. No status page was found on surgepix.ai or in the docs, and status.surgepix.ai doesn't resolve (0). With no history to read, the incident line takes the floor (5). No rate limits with numbers are documented. The terms say "non-human continuous or frequent query operations" lead to automatic account suspension, with no threshold given (0). No 429 or backoff guidance. The skill scripts poll every 2 seconds for up to 5 to 15 minutes, but that is client code, not a documented policy (0). No SLA, and the terms disclaim uninterrupted service (0). The API isn't labelled beta or preview, and API keys have been self-serve since 2 June 2026 per the changelog (10). 0+5+0+0+0+10 = 15.

Iris-3B 52

Apache-2.0 for code and weights, with a NOTICE file, a third-party licence table and the full training and distillation recipe. The training images and the Turbo prompt sets are not released or named in the README or card (25 of 30). Self-hosted, so inputs stay on the owner's hardware, but we found no statement saying so and speridlabs.com has no privacy policy (15 of 30). No deprecation policy, version tags or pinned revisions (2 of 20). No telemetry calls found in the inference code. Training defaults to report_to: wandb, which sends run config and validation images to Weights & Biases, can be set to none, and is not mentioned in the README (8 of 20).

SurgePix API 28

The editorial half. A closed service whose terms name no company or address. They restrict any commercial use of what the service produces without SurgePix's permission, and contain clauses from an unrelated product (a "big data" query platform and "non-resume content"). The skills are MIT. Unclear terms, with a little credit for the open skills (5). The privacy policy is dated both 10 June 2026 and 1 July 2022, describes WeChat and phone registration and "domestic servers", and gives no retention periods. The API reference says uploads go to Tencent Cloud storage in Seoul as public URLs. No DPA. The statements don't agree (5). No deprecation policy or notices, and the terms let changes take effect on publication without notice (0). Tencent Cloud Object Storage in ap-seoul is named in the API reference, and Alipay and WeChat Pay in the privacy policy. No subprocessor list or stated data location (4). 5+5+0+4 = 14.

Iris-3B 47

Read for a model the owner runs through a CLI and a Python package. No server, OpenAPI or other machine-readable contract. The CLI uses argparse with typed flags and help text, and the Python functions carry type hints. The demo Spaces' Gradio API lists typed endpoints, but Speridlabs doesn't document it (8 of 25). No llms.txt. The README and model card are Markdown in each repository (5 of 10). The card and README state limits (unreliable rendered text and object counts, about one megapixel, depth relative not metric, Turbo's fixed schedule and lower seed variety) but not intended or out-of-scope uses (14 of 20). Flags are typed, with the size presets as an enum only in the Spaces. Height and width take any integer, though Turbo was distilled at seven sizes (9 of 15). Commands for every model and a tips table, but no documented errors (7 of 15). No versioning or changelog beyond dated commits (4 of 15).

SurgePix API 40

No OpenAPI or other machine-readable contract was found (0). surgepix.ai/llms.txt returns 404. The docs are HTML pages, but the 15 SKILL.md files in the MIT repository are Markdown written for agents, so half (5). The API reference states the purpose of each of its five operations and the async and session model. The SKILL.md files add trigger phrases and explicit do-not-use routing between skills, but 11 of the task endpoints the skills call appear only in their scripts, with no reference (10). Parameter tables give type and required for the documented operations, with no enums or limits, and the undocumented endpoints have no schema at all (7). JSON examples for every documented operation and an error table with 401, 404 and error.file.sensitive_content (10). A dated public changelog from 11 March 2026, but no version in the API path and no API-specific change notes (8). 0+5+10+7+10+8 = 40.