# Deepgram Voice Agent API vs Inworld Realtime API (slim)

> Deepgram Voice Agent scores 72.2 (BB) to Inworld Realtime's 60.1 (C) for voice agents. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/deepgram-voice-agent-vs-inworld-realtime.md (~5,550 tokens) · this version ~2,830 tokens · JSON https://www.anchorterminal.com/compare/deepgram-voice-agent-vs-inworld-realtime.json · canonical https://www.anchorterminal.com/compare/deepgram-voice-agent-vs-inworld-realtime
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-07
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Deepgram Voice Agent API scores 72.2 (BB) on agent readiness against Inworld Realtime API's 60.1 (C), and leads in 5 of 7 scored categories. Both do voice agents.

- Deepgram Voice Agent API: BB 72.2, rank #136 of 1510 · https://www.anchorterminal.com/tools/deepgram-voice-agent.min.md
- Inworld Realtime API: C 60.1, rank #671 of 1510 · https://www.anchorterminal.com/tools/inworld-realtime.min.md
- Deepgram Voice Agent API, good for Developers who already run telephony (Twilio, Amazon Connect, Genesys, AudioCodes) and want a cheap managed pipeline with their choice of LLM. Ahead on Reliability, 75 against 62; Schema & documentation, 90 against 60; Security & auth, 71 against 58; Maintenance & community, 85 against 71; Transparency & trust, 78 against 63. Also Agent-ready, a grade of BB or better; A hosted endpoint, with nothing to install; Runs on your own machine; Free to start without a card.
- Inworld Realtime API, good for Teams moving off OpenAI's Realtime API who want a choice of LLM and Inworld voices.

| Category | Deepgram Voice Agent API | Inworld Realtime API |
| --- | --- | --- |
| Reliability (16%) | 75 | 62 |
| Performance (10%) | pending | pending |
| Schema & documentation (13%) | 90 | 60 |
| Agent ergonomics (13%) | 67 | 68 |
| Security & auth (14%) | 71 | 58 |
| Payments & pricing (10%) | 40 | 40 |
| Task success (10%) | pending | pending |
| Maintenance & community (7%) | 85 | 71 |
| Transparency & trust (7%) | 78 | 63 |

| Fact (where they differ) | Deepgram Voice Agent API | Inworld Realtime API |
| --- | --- | --- |
| Vendor | Deepgram | Inworld AI |
| Hosted endpoint | `https://agent.deepgram.com/v1/agent` | no (local only) |
| Transports | HTTP, Streamable HTTP, stdio, SSE (legacy) | websocket, HTTP |
| Price for voice agents | not published | $25 per 1M characters |
| Licence | MIT (SDKs) | none |
| Tools exposed | none | 0 |
| Last release | 2026-10-28 | 2026-10-16 |
| Terms last updated | 2026-08-06 | 2025-06-11 |
| Privacy policy last updated | 2021-10-26 | 2026-01-18 |
| Customer content may train models | yes, with an opt-out | not found in the text |
| Terms restrict benchmarking | yes | not found in the text |
| Terms or service can change without notice | yes | not found in the text |
| Arbitration or class-action waiver | yes | not found in the text |
| Popularity | 468 stars, 1.1M npm/wk, 805k PyPI/wk | none |
| Agent reviews | 3/5 (2) | none |

Why the edge:
- Schema & documentation, Deepgram Voice Agent API +30: Deepgram Voice Agent API 90 (Public OpenAPI and an AsyncAPI file that describes the agent socket messages (25). llms.txt with a Markdown copy of each page, and a keyless docs MCP server (10). Every client message and server event has its own page saying what it's for, from `Settings` to `ForceEndTurn` and `FunctionCallRequest` (17 of 20). The AsyncAPI types the Settings message with enums and required fields, though provider `endpoint.headers` are free-form (13 of 15). Examples in every language guide and a Voice Agent errors and warnings page, which we read only in outline (11 of 15). A dated changelog that also publishes corrections, such as the token TTL field name on 18 September 2026 (14 of 15).); Inworld Realtime API 60 (No AsyncAPI or other machine-readable file for the socket, and the spec paths redirect (0). llms.txt, Markdown twins and a docs MCP server (10). The overview, OpenAI migration guide, `providerData` reference and the billing page's account of how a session is counted say what to use when (16 of 20). The reference lists every client and server event with field types and required marks, as text rather than a schema (12 of 15). JavaScript and Python example projects and an error event with `type`, `code` and `param`, without a list of codes (9 of 15). Dated release notes with affected clients and required actions, but no API version beyond `/v1` (13 of 15). Total 60.)
- Transparency & trust, Deepgram Voice Agent API +15: Deepgram Voice Agent API 78 (The service is closed with clear terms, and the official SDKs are MIT (15 of 30, read as deepgram-stt and deepgram-tts are). The docs say audio and transcripts are kept for model improvement unless `mip_opt_out` is set and describe EU, Australian and Indian endpoints, while the privacy policy (last updated 26 October 2021, re-read on 1 November 2026) says data is stored on US servers and doesn't mention the Model Improvement Program, so the statements disagree (15 of 30). The changelog dates changes and corrections, but we didn't find a published deprecation policy for the agent API (10 of 20). A subprocessor page and EU, India and Australia endpoints are published (20). 15 + 15 + 10 + 20 = 60. The 1 October check gave the licence line 18 and the data-handling line 22, where Deepgram's other listings read the same pages as 15 and 15.); Inworld Realtime API 63 (Closed service with clear terms naming Theai, Inc. dba Inworld AI (15). The ZDR page says what a session keeps and that third-party LLM requests aren't covered, but the terms and privacy notice disagree on training (12 of 30). Release notes flag potentially breaking changes with the clients affected and what to do, without stating a notice period (14 of 20). US headquarters with EU and India residency on Enterprise, and no sub-processor list in the docs, while sessions may route to third-party LLM providers (10 of 20). Total 51.)
- Maintenance & community, Deepgram Voice Agent API +14: Deepgram Voice Agent API 85 (Changelog entries to 27 October 2026 and the Python and JavaScript SDKs released on 28 October (Python 7.16.0, JavaScript 5.17.0), both adding Voice Agent Think Live messages (30). 34 dated changelog entries in the 90 days from 3 August, several on the Voice Agent API, and eight Python SDK tags from 7.9.0 on 14 September to 7.16.0 on 28 October (20). A public changelog that issues corrections, such as the 8 October correction of the over-limit `Configure` behaviour, plus support channels we didn't test (12 of 15 for a closed service). Official SDKs current in Python, JavaScript, Java, Go, .NET and Rust, with Voice Agent support for custom Think-provider messages from 8 October (15). MIT SDKs with current releases (8 of 10). 30 + 20 + 12 + 15 + 8 = 85.); Inworld Realtime API 71 (Release note of 16 October 2026 (`cache_transcript`), 21 days ago (30). Three dated Realtime entries in the 90 days from 8 August, on 14 September, 23 September and 16 October (20). Discord and support@inworld.ai, with no evidence of answers read (8 of 15). Works with OpenAI's Realtime clients and the CLI is current at 1.5.0, but there is no Inworld Realtime SDK (8 of 15). CLI on npm under MIT, CI not read (5 of 10). Total 71.)
- Reliability, Deepgram Voice Agent API +13: Deepgram Voice Agent API 75 (Statuspage at status.deepgram.com with component history, a Voice Agent API component and an RSS feed (20). Nineteen incidents in the 90 days from 3 August, and every one is partial by its own text (one datacenter, one region or endpoint, one model or some requests), read as this run reads partial incidents and as deepgram-stt and deepgram-tts are read on the same page. The only one that names the Voice Agent API is 29 October, when authentication at one US datacenter failed some API requests and Voice Agent sessions with 5xx errors from 15:17 to 15:45 UTC. The STT and TTS incidents the pipeline sits on include Flux TTS errors on the global endpoint only from 09:45 to 13:44 UTC on 25 September, some batch STT requests slow or timing out from 19:27 to 21:53 UTC on 22 October and 5xx errors on some requests at one datacenter from 16:00 to 17:09 UTC on 22 October. The 7 July Flux errors and the 21 July Gemini routing failure, read on 1 October with others as several majors, have left the window (cause time). Minor only (20). Concurrency is published, 45 sockets on pay as you go, 60 on Growth in North America and from 100 on Enterprise (15). Over-limit requests get a 429 and the docs recommend exponential backoff, with no Retry-After header and no idempotency guidance for function calls (10 of 15). The pricing page no longer names an uptime level, and we found no SLA terms (0). The Voice Agent API is generally available (10). 20 + 20 + 15 + 10 + 0 + 10 = 75.); Inworld Realtime API 62 (Hosted reading of the checklist. status.inworld.ai is a public status page (20), with history we couldn't read, as on inworld-tts (5). Concurrent sessions are published per plan, 10 on On-Demand up to 1,000 on Growth (15). The rate limits page gives exponential backoff with jitter and Retry-After on longer windows, and sessions emit `rate_limits.updated`, with nothing on reconnecting a dropped session (12 of 15). An SLA appears only in the Enterprise column, with no terms (0). No preview or beta label on the Realtime API (10). 20 + 5 + 15 + 12 + 0 + 10 = 62.)
- Security & auth, Deepgram Voice Agent API +13: Deepgram Voice Agent API 71 (API keys carry an owner, admin or member role, can be set to expire, can be created through the API and can be tagged, and browsers get 30-second JWTs from the token endpoint (30). Member keys limit what a leaked key can do, and the function-call hold keeps the agent from firing irreversible tools on a speculative turn, but there's no read-only agent mode (14 of 20). The agent feeds caller speech to an LLM and we found no prompt-injection guidance (5 of 15). Requests are tagged per key and each session reports latency and history, but we found no audit log of account actions (10 of 15). The compliance page states SOC 2 Type 1 and Type 2, HIPAA and PCI. deepgram.com now serves a valid security.txt (Contact security@deepgram.com, Expires 30 September 2027), not found on 30 September. No disclosure policy, bug bounty or public advisories found (12 of 20, the same reading as deepgram-stt and deepgram-tts). 30 + 14 + 5 + 10 + 12 = 71.); Inworld Realtime API 58 (Workspace API keys sent as Basic credentials, deleted to revoke, with separate keys per environment advised. The CLI mints read-only keys unless `--write` is given, but write toggles cover only the Voice and Router APIs, so no key is limited to one product. Browsers use short-lived tokens minted server side. Session IDs, not secrets, go in the URL (25 of 30). Tools run in the client, so the developer approves side effects, with no approval step in the API (10 of 20). The agent listens to callers and passes their speech to an LLM, and no prompt-injection guidance was found (5 of 15). Usage per key and a Performance page with errors and throttling (8 of 15). SOC 2 Type II and HIPAA claimed and a vulnerability report form, with no security.txt or bug bounty, per the 4 November reading (10 of 20). Total 58. Zero data retention covers session content from the Developer plan up, but not requests routed to third-party LLM providers.)