Mistral Embed and Codestral Embed by Mistral AI

# Fix list: Mistral Embed and Codestral Embed

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/mistral-embeddings, the October 2026 research run, assessed 1 October 2026. Grade C, 58.2 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 Mistral Embed and Codestral Embed: 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. Reliability, 38 out of 100, up to 12.4 more on the total

Why it scored 38: status.mistral.ai runs on Rootly with an Embedding API component and 90 days of uptime bars (20). The history shows two incidents titled Embedding API Degraded, opened on 12 August 2026 at 16:29 UTC and 27 August 2026 at 18:04 UTC, and the Embedding API's 90-day uptime reads 94.36 per cent, the lowest of the 15 components and about five days of lost uptime, so worse than several majors (0). Limits are shown per workspace in the admin panel, and the usage-limits page publishes no numbers for embeddings (0 of 15). An error glossary explains each status code, per the Mistral AI API listing's check, and we found no Retry-After or backoff guidance (8 of 15). No SLA found (0). Both embedding models are GA, not Labs or preview (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.

## 2. Security & auth, 45 out of 100, up to 9.6 more on the total

Why it scored 45: Plain API keys per workspace, revocable in the console, no endpoint scopes found (20). The same key reaches files, fine-tuning, agents and batch jobs, including deletes, with no way to limit it to embeddings (10 of 20). Returns vectors only (10). No per-key log or audit trail found in the docs we read (0 of 15). security.txt is valid, as checked for the Mistral AI API listing, and we couldn't confirm a bug bounty or certifications in this run (5 of 20).

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.

## 3. 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-token prices public, $0.10 per million for mistral-embed and $0.15 for codestral-embed, batch at half price (20). A free tier with included monthly usage and no card, though it needs a phone number and its data may be used for training (20). A person signs up in a browser and verifies a phone (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.

## 4. Maintenance & community, 40 out of 100, up to 5.3 more on the total

Why it scored 40: No new embedding model since codestral-embed on 28 May 2025, the only changelog entry that mentions embeddings, and mistral-embed dates from 11 December 2023 (0). Dated changelog entries on 16 July, 31 August, 28 September and 29 September 2026, all for other models (20). client-python has 29 open issues, and the twelve newest, from 30 June to 26 September 2026, show no maintainer reply in the issue list, including an open report that the pinned cryptography version carries security alerts (5 of 25). Official SDKs, mistralai on PyPI and @mistralai/mistralai on npm, release dates not checked in this run (10 of 15). SDKs generated from the OpenAPI spec, CI not checked (5 of 10).

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.

## 5. Agent ergonomics, 78 out of 100, up to 3.6 more on the total

Why it scored 78: codestral-embed takes output_dimension up to 3072 and int8, uint8, binary or ubinary output, but mistral-embed is fixed at 1024 floats (20 of 25). Batched input lists and the batch endpoint, no truncation switch documented (12 of 20). The error glossary gives a meaning and fix per status code (16 of 20). Stateless calls, safe to retry, but no retry guidance found (15 of 20). Two required parameters, official SDKs in Python and TypeScript (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.

## 6. Schema & documentation, 89 out of 100, up to 1.8 more on the total

Why it scored 89: Public OpenAPI document at docs.mistral.ai/openapi.yaml covering /v1/embeddings (25). llms.txt at docs.mistral.ai (10). Separate text and code embedding pages say which model fits which job and how to cut codestral-embed's dimensions (14 of 20). model and input required, output_dimension and output_dtype typed, with the dtype values float, int8, uint8, binary and ubinary on codestral-embed (13 of 15). SDK and curl examples on the embedding pages and an error glossary with a fix per status (12 of 15). Dated changelog and dated model ids such as mistral-embed-2312 and codestral-embed-2505 (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.

## 7. Transparency & trust, 81 out of 100, up to 1.7 more on the total

Made of editorial 65, provenance 96.

Why it scored 81: Closed service under commercial terms with a French legal entity, SDKs Apache-2.0 (15). Data sent to Labs and preview models is used for training under the terms effective 25 September 2026, neither embedding model is one, free-tier data may train models, abuse logs are kept 30 days unless zero retention is bought, and the paid default isn't spelt out (15 of 30). A model lifecycle page gives notice periods per stage, six months for GA models, per the Mistral AI API listing's check (20). EU and US regional endpoints at 1.1 times the price say where data can be processed, subprocessor list not checked (15 of 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):

- Domain age: mistral.ai, registered 2019-05-15 (7 years) (11 of 15)

## 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.

- How long each August 2026 Embedding API degradation lasted. The history lists the start times only, and the 94.36 per cent uptime implies days rather than hours.
- Rate limits for the embedding models, which are only visible in the admin panel.
- Whether the paid tier trains on embedding inputs. The terms are explicit only for Labs, preview and free-tier use.
- Certifications and a bug bounty, which we couldn't confirm in this run.

## Weaknesses

- Embedding API uptime of 94.36 per cent over 90 days on Mistral's status page, with incidents on 12 and 27 August 2026
- 8k context on both models, and text or code only
- mistral-embed dates from December 2023 with fixed 1024-dimension float output, and nothing new since May 2025
- No reranker, no published rate limits and no language list for the embedding models
- Free-tier data may be used for training

## 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.

- Use codestral-embed whenever you want smaller or binary vectors. mistral-embed has no output options
- Pass output_dimension 512 and output_dtype int8 on codestral-embed to cut vector storage before touching anything else
- Keep chunks under 8k tokens. There's no long-context embedding model on this API
- Check status.mistral.ai before a big index job and retry with backoff, since the Embedding API had two degradations in August 2026
- Pin dated model ids (mistral-embed-2312, codestral-embed-2505) so an alias move can't change your vectors

## What the review panel asked for

- Publish embedding limits
- Say which parameters each model accepts
- Publish embedding rate limits

## 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.