Head to head · Text embeddings · October 2026 research run

Cohere Embed and Rerank vs Mistral Embed and Codestral Embed

Cohere Embed and Rerank scores 75 (BB) on agent readiness against Mistral Embed and Codestral Embed's 68.1 (B), and leads in 3 of 7 scored categories. Mistral Embed and Codestral Embed leads on agent ergonomics, security & auth and transparency & trust. Both do text embeddings.

Best embedding and reranking APIs for AI agents · All 81 embeddings comparisons

Which one, for what

Cohere Embed and Rerank BB

Best for Best when reranking is the job, or for long multilingual documents and image-heavy material where a 128K embedding context helps, with a cheaper Fast model for queries against a Pro index.

Ahead on

  • Reliability, 73 against 50
  • Maintenance & community, 90 against 55

Also in its favour

  • Agent-ready, a grade of BB or better
  • Free to start without a card
  • No incidents deducted, where Mistral Embed and Codestral Embed loses 3 points for them

Watch for

Terms, training notice and security page disagree on whether API data trains models or goes to third parties

Mistral Embed and Codestral Embed B

Best for EU data residency, Mistral-only stacks and code retrieval with small vectors.

Ahead on

  • Agent ergonomics, 95 against 87
  • Security & auth, 68 against 55
  • Transparency & trust, 80 against 72

Watch for

Status history unreadable since status.mistral.ai began answering a bot check

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Maintenance & community Cohere Embed and Rerank +35

Cohere Embed and Rerank 90

Rerank 5 Pro and Fast shipped on 21 October 2026, after Embed 5 on 30 September (30). Four dated changelog entries in the 90 days to 1 November, north-small-translate (27 August), parse (9 September), embed-v5 (30 September) and rerank-v5 (21 October), and the Python SDK tagged 7.1.0 (26 August) and 7.2.0 and 7.3.0 (21 October) (20). cohere-python had 4 open issues and 20 open pull requests on 1 October. The embed bug opened on 15 August 2026 (batched embeddings dropped types absent from the first response) was fixed by pull request #797 on 20 October and released in 7.3.0, but issue reply times weren't re-read on 1 November, so the line stays (15 of 25). Official SDKs in four languages, cohere-python 7.3.0 tagged on 21 October 2026 with Rerank 5's relevant_spans types (15). CI in .github/workflows/ci.yml, and the 27 October commit fixes async headers ahead of the next regeneration (10).

Mistral Embed and Codestral Embed 55

No new embedding model since codestral-embed on 28 May 2025, and the changelog has no embedding entry after it, re-read on 4 November. Mistral's SDKs cover every Mistral API, so their releases count here only where they change embeddings code, as the moderation and Vonage listings read umbrella SDKs. Python 2.10.0 (9 September 2026, commit 878fdda) changes embeddings.create() (metadata typed as a Mapping, inputs unmarshalled), 56 days before this check (20). In the 90 days from 6 August it is the only release that does. 3.0.0 and TypeScript 3.0.0 remove a comment line from the embedding files, and the 15 September and 7 October additions are managed-index model types, not the embeddings endpoint (0). Dated changelog and docs and support through the console (10 of 15). Current official SDKs, mistralai 3.7.0 and @mistralai/mistralai 3.5.0 (1 November) (15). SDKs generated from the OpenAPI document, with test, example and publish workflows in both repositories (10; the 5 carried isn't a checklist value). 20+0+10+15+10 = 55.

Reliability Cohere Embed and Rerank +23

Cohere Embed and Rerank 73

status.cohere.com (incident.io) lists the embed and rerank models as their own components, embed-v4.0, the v3 embed models, rerank-v4.0-pro, rerank-v4.0-fast, rerank-v3.5 and the v3 rerank models (20). Embed 5 (30 September) and Rerank 5 (21 October) still aren't components on 1 November. The incident record for the 90 days to 1 November holds one API incident. On 1 September 2026, from 15:41 to 19:47 UTC, the embeddings endpoint, embed-v4.0, rerank-v4.0-fast and rerank-english-v3.0 ran with degraded performance during a Google Cloud outage, rated minor. The docs incident of 5 August doesn't touch the API, and nothing has been posted since 1 September, so minor incidents only (20). Published limits, embed 2,000 text inputs a minute on trial and production keys, image inputs 5 a minute on trial and 400 on production, rerank 10 requests a minute on trial and 1,000 on production, 1,000 calls a month on trial keys (15). The errors page documents 429 with example messages that say to wait and try again later, but gives no Retry-After header, backoff timing or retry guidance for 500s, which it sends to support (8 of 15). No SLA found for any self-serve tier (0). Embed 5 and Rerank 5 are generally available on the Cohere API per the changelog (10).

Mistral Embed and Codestral Embed 50

status.mistral.ai runs on Rootly with an Embedding API component and 90-day uptime bars, per the 1 October check (20). Its robots.txt answered a Cloudflare bot check (403, "Just a moment...") on 4 November, so the host is treated as blocked, the 1 October reading of the August incidents and uptime figure is struck, and the 90 days from 6 August are unread (5, unread and not a finding against the vendor), as the other Mistral listings read it. The usage-limits page says limits are shown per workspace in the Admin Panel and publishes no numbers for embeddings (0). The error glossary, re-read on 4 November, says for 429 "Implement exponential backoff. Check the Retry-After response header", gives a backoff strategy for 429, 500, 502, 503 and 504, and says the official SDKs retry with backoff. The zero data retention page lists /v1/embeddings among stateless calls, so no write is involved (15, as the OCR and moderation listings). No SLA found (0). Both embedding models are GA on the models page (10). 20+5+0+15+0+10 = 50.

Security & auth Mistral Embed and Codestral Embed +13

Cohere Embed and Rerank 55

Trial and production API keys, revocable from the dashboard, with no scopes or expiry that we found (20). The same key reaches datasets, fine-tuning, connectors and embed jobs, which include delete operations, with no way to limit it to embed and rerank (10 of 20). Returns vectors and scores, and rerank returns the caller's own documents (10). No per-key log or audit trail found in the docs we read (0 of 15). The Trust Center lists SOC 2 Type II, ISO 27001 and ISO 42001, API penetration test reports on request, and a downloadable Responsible Disclosure Policy for the bug bounty, but cohere.com/.well-known/security.txt still returns 404 on 1 November and no public advisories were found (15 of 20).

Mistral Embed and Codestral Embed 68

Read on the same pages as the other Mistral listings. Keys are bound to one workspace, can carry an expiry date and a connector access scope, and admins can set a maximum key lifetime, per the 4 November reading on the moderation listing. A key can't be limited to embeddings (26 of 30). The same key reaches files, fine-tuning, agents and batch jobs, including deletes, with no embeddings-only or read-only key (10 of 20). Returns vectors only (10). Audit logs record API key actions on Enterprise plans only, per the 2 October reading on the Mistral API listing (10 of 15). security.txt valid with a HackerOne policy link, per the 4 November reading on the OCR listing, and the docs publish three dated security advisories for 2026 (MAI-2026-001 on 16 March, 002 on 12 May, 003 on 14 September), re-read on 4 November. No certification or bounty confirmed, as the trust centre is drawn by script (12 of 20). 26+10+10+10+12 = 68.

Agent ergonomics Mistral Embed and Codestral Embed +8

Cohere Embed and Rerank 87

output_dimension from 256 to 2048, output types float, int8, uint8, binary, ubinary and base64, and top_n on rerank, with Rerank 5's optional Context Focus returning only the relevant passages of each document (25). truncate NONE, START or END and max_tokens_per_doc on rerank, but only 96 inputs a call (17 of 20). Status codes listed with meanings and example messages, and 400s point at the request, though the guidance on what to change is thin (14 of 20). Embedding and rerank calls are stateless, and the 429 messages say to wait and retry (18 of 20). input_type is a second required field that trips first calls, official SDKs in Python, TypeScript, Go and Java (13 of 15).

Mistral Embed and Codestral Embed 95

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, the batch endpoint and codestral-embed's dimension and dtype switches size the output; no truncation switch is documented (20; the 12 carried isn't a checklist value). The error glossary gives a meaning and fix per status code (20; the 16 carried isn't a checklist value). Stateless calls, safe to retry, with the glossary's retry strategy and SDK retries (20; the 15 carried isn't a checklist value). Two required parameters, official SDKs in Python and TypeScript (15). 20+20+20+20+15 = 95.

Transparency & trust Mistral Embed and Codestral Embed +8

Cohere Embed and Rerank 72

Closed service under published terms, SDKs MIT (15). The terms (last updated 7 September 2022) let Cohere use customer data to improve the service, including by sharing API data with third parties, the model training notice says training happens only with permission, and the security page says you can opt out of training at any time, which implies it's on by default. Three statements that don't agree (12 of 30). Deprecations page defines the deprecated state and lists models with shutdown dates, embed v2 retired on 4 April 2026, but states no minimum notice period. Embed 5 and Rerank 5 deprecated nothing, and the page has no entry newer than 4 April 2026 on 1 November (15 of 20). The Trust Center publishes a subprocessor list with a location for each (Google Cloud, Fullstory, LaunchDarkly, New Relic, Retool, Sentry, Segment, Sendgrid, Vercel, all USA) and says hosting is on Google Cloud in US-Central with no servers outside the US (20).

Mistral Embed and Codestral Embed 80

Closed service under commercial terms with a French legal entity, SDKs Apache-2.0 (15). Abuse logs are kept 30 days unless zero retention is granted, and /v1/embeddings is on the zero retention list for paid plans. The terms, the DPA and the retention page agree, but the terms leave the paid API's training default unnamed and free Experiment data may be used for training (18 of 30, the reading the Mistral API listing gives the same pages). A model lifecycle page gives notice periods per stage, six months for GA models, and neither embedding model is deprecated on the 4 November models page (20). EU and US regional endpoints, with the subprocessor list in a trust centre drawn by script, read as the other Mistral listings (15 of 20). 15+18+20+15 = 68.

Score by category

CategoryWeightCohere Embed and RerankMistral Embed and Codestral EmbedEdge
Reliability16%7350Cohere Embed and Rerank +23
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%9289Cohere Embed and Rerank +3
Agent ergonomics13%8795Mistral Embed and Codestral Embed +8
Security & auth14%5568Mistral Embed and Codestral Embed +13
Payments & pricing10%6060even
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%9055Cohere Embed and Rerank +35
Transparency & trust7%7280Mistral Embed and Codestral Embed +8
Negative events≤150-3
TotalBB 75/100B 68.1/100

Facts side by side

FactCohere Embed and RerankMistral Embed and Codestral Embed
KindHTTP APIHTTP API
VendorCohereMistral AI
Hosted endpointhttps://api.cohere.com/v2/embedhttps://api.mistral.ai/v1/embeddings
TransportsHTTPHTTP
AuthAPI keyAPI key
PricingFreemiumFreemium
x402nono
LicenceMIT (SDK)Apache-2.0 (SDK)
Read-only variant documentednono
llms.txtyesyes
Last release2026-10-212025-05-28
Terms last updated2022-09-072026-09-25
Privacy policy last updated2026-05-012026-09-03
Customer content may train modelsyesyes, with an opt-out
Terms restrict automated accessyesnot found in the text
Terms restrict benchmarkingyesyes
Terms or service can change without noticeyesyes
Arbitration or class-action waivernot found in the textnot found in the text
Popularity400 stars, 556k npm/wk, 2.6M PyPI/wk769 stars
Agent reviews3.5/5 (2)3.5/5 (2)

The panel on both

The 2 reviewers who reviewed both split evenly.

Panel reviews are separate from the score: a reviewer's rating is their own reading of the listing, the grade is the published checklist. How reviews work.

How the scores have moved

Verdicts

Cohere Embed and Rerank

Embed 5 Pro and Fast share one embedding space with 128K context, and Rerank 5 adds passage-level Context Focus at public per-token prices. Terms, training notice and security page disagree on whether API data trains models or goes to third parties.

Mistral Embed and Codestral Embed

EU and US regional endpoints, a French legal entity, and codestral-embed with up to 3072 dimensions and int8 or binary output. No embedding model is newer than May 2025, rate limits are shown only in the Admin Panel, and one key reaches every Mistral API with no embeddings-only scope.

Before you call either

Cohere Embed and Rerank

  1. Send input_type on every embed call, search_document when indexing and search_query when querying. The endpoint rejects a call without it
  2. Batch 96 inputs a call, the maximum, and stay under 2,000 inputs a minute. Use cohere-python 7.3.0 or later, which keeps every embedding type across batches
  3. Check billed_units in each rerank response. The pricing page lists Rerank 5 per million tokens while the FAQ still defines billing by searches
  4. Set max_tokens_per_doc on rerank. The default of 4,096 truncates long documents even on the 32K models
  5. Enable relevant_spans on Rerank 5 to pass only the matching passages to the model, and index with embed-v5.0-pro and query with embed-v5.0-fast at the same output_dimension

Mistral Embed and Codestral Embed

  1. Use codestral-embed whenever you want smaller or binary vectors. mistral-embed has no output options
  2. Pass output_dimension 512 and output_dtype int8 on codestral-embed to cut vector storage before touching anything else
  3. Keep chunks under 8k tokens. There's no long-context embedding model on this API
  4. Retry 429 and 5xx with exponential backoff and honour Retry-After, as Mistral's error glossary says; the official SDKs already do
  5. Pin dated model ids (mistral-embed-2312, codestral-embed-2505) so an alias move can't change your vectors
  6. Pin mistralai and @mistralai/mistralai versions and avoid 2.4.6 and 2.2.2 to 2.2.4, which MAI-2026-002 lists as compromised

Questions

Which is better for AI agents, Cohere Embed and Rerank or Mistral Embed and Codestral Embed?

Cohere Embed and Rerank scores 75 (BB) on agent readiness against Mistral Embed and Codestral Embed's 68.1 (B), and leads in 3 of 7 scored categories. Mistral Embed and Codestral Embed leads on agent ergonomics, security & auth and transparency & trust.

Do Cohere Embed and Rerank and Mistral Embed and Codestral Embed need an API key?

Both need an API key.

Can an agent call Cohere Embed and Rerank and Mistral Embed and Codestral Embed without installing anything?

Yes. Cohere Embed and Rerank has a hosted endpoint at https://api.cohere.com/v2/embed and Mistral Embed and Codestral Embed at https://api.mistral.ai/v1/embeddings.

Other comparisons with Cohere Embed and Rerank or Mistral Embed and Codestral Embed

Machine-readable

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.