Deepgram Speech-to-Text (Nova-3, Flux) by Deepgram

# Fix list: Deepgram Speech-to-Text (Nova-3, Flux)

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/deepgram-stt, the October 2026 research run, assessed 1 October 2026. Grade BB, 70.6 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Deepgram Speech-to-Text (Nova-3, Flux): work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Payments & pricing, 40 out of 100, up to 7.5 more on the total

Why it scored 40: No x402, MPP or L402 (0). Per-minute prices for every model and add-on published without a login (20). $200 credit with no card (20). A person signs up in a browser to get the first key (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 2. Reliability, 65 out of 100, up to 7 more on the total

Why it scored 65: Statuspage at status.deepgram.com with component history (20). Seven incidents between 7 July and 30 September 2026. Two count as major, elevated Flux WebSocket errors for 2 hours 24 minutes on 7 July and batch 400 and then 5xx errors for about 2.5 hours on 28 July, with monitoring until the next day. The rest were under 70 minutes or hit only the Voice Agent API (5, between the rubric's 10 for one major and 0 for several). Concurrency limits per project published with numbers, 50 pre-recorded and 150 streaming on pay as you go (15). Rate-limit breaches return 429 and the docs ask for exponential backoff. Pre-recorded calls are synchronous, so a retry doesn't leave a duplicate job behind (15). No SLA found for self-serve plans (0). Nova-3 and Flux are GA (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 3. Security & auth, 65 out of 100, up to 6.1 more on the total

Why it scored 65: Model reading of the checklist, with training and retention in place of least-privilege and injection lines. Keys belong to a project, carry a role and can expire on a date or after a set duration. Browser JWTs live 30 seconds and travel in a header (30). Requests may be kept for model training under the Model Improvement Program unless each request sets `mip_opt_out=true`, which any account can do (10). Opted-out requests are kept only for processing, so zero retention is self-serve but not the default (10). Per-request logs and usage data in the console and API, no audit log of key actions found (10). SOC 2 Type II report on request (5). No security.txt, disclosure policy, bug bounty or public advisories found (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 4. Agent ergonomics, 75 out of 100, up to 4.1 more on the total

Why it scored 75: API reading of the checklist. Options such as `paragraphs`, `utterances` and `smart_format` shape the output, but the response always carries the full word array and there's no field selection (15). The request history endpoint pages and filters by date and status (20). Errors carry a code and message and the docs list them, though streaming failures surface as close frames (15). No idempotency key, and a retried pre-recorded call is billed again (10). Only the audio is required. Official SDKs in six languages, including Python, JavaScript, Go, .NET and Java (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 5. Transparency & trust, 75 out of 100, up to 2.2 more on the total

Made of editorial 60, provenance 90.

Why it scored 75: Closed service with clear terms (15). The privacy policy was last updated on 26 October 2021, says data is stored on US servers and doesn't mention the Model Improvement Program, while the docs describe training by default and EU, Australian and Indian endpoints, so the statements disagree (15). The changelog flags breaking corrections, such as two on 18 September 2026, but we found no deprecation policy (10). Sub-processor list at deepgram.com/privacy/subprocessors and four regional endpoints (20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- security.txt: not found (0 of 10)

## 6. Maintenance & community, 80 out of 100, up to 1.8 more on the total

Why it scored 80: Changelog entries on 24 and 25 September 2026 for Nova-3 languages and Flux numerals, and SDK releases on 14 September (30). More than 15 dated entries in the last 90 days (20). Public changelog and community support channels (10). Python 7.9.0, JavaScript 5.11.0 and Java 0.10.0 released on 2026-09-14 (15). SDK CI not checked in this run (5).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## 7. Schema & documentation, 95 out of 100, up to 0.8 more on the total

Why it scored 95: OpenAPI file at developers.deepgram.com/openapi.json and an AsyncAPI file for the sockets (25). llms.txt (10). Reference pages state what each option does, and the docs say to use Flux for live agents and Nova-3 for files, though few options say when not to use them (15). Query parameters are typed with enums for `model`, `language` and `encoding` (15). Examples on every endpoint and a documented error format (15). Versioned `/v1` and `/v2` paths and a dated changelog (15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- Whether an SLA exists for any self-serve plan
- Whether the SDK repositories run CI on the default branch
- When the privacy policy will reflect the Model Improvement Program and the regional endpoints

## Weaknesses

- Training on audio is the default and the opt-out is a per-request flag
- Two incidents over 2 hours in July 2026, on Flux streaming and batch
- No SLA published for self-serve plans
- The privacy policy dates from October 2021 and doesn't mention the Model Improvement Program
- Streaming prices are promotional and may rise to the regular rate

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Add `mip_opt_out=true` to every request that carries customer audio
- Use Flux (`flux-general-en`) on `/v2/listen` for live agents and Nova-3 for files
- Back off exponentially on 429. The concurrency limit is per project
- Pass `callback` for long files so the request doesn't hit the 10-minute processing timeout
- Mint keys with an expiry for short-lived jobs

## What the review panel asked for

- Publish promotion end date
- Publish a self-serve SLA
- Add an idempotency key

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.