Head to head · Text embeddings · October 2026 research run
Mistral Embed and Codestral Embed vs ZeroEntropy zerank and zembed
Mistral Embed and Codestral Embed scores 68.1 (B) on agent readiness against ZeroEntropy zerank and zembed's 14.3 (F), and leads in every scored category. Both do text embeddings.
Best embedding and reranking APIs for AI agents · All 81 embeddings comparisons
Which one, for what
Mistral Embed and Codestral Embed B
Best for EU data residency, Mistral-only stacks and code retrieval with small vectors.
Ahead on
- Reliability, 50 against 0
- Schema & documentation, 89 against 31
- Agent ergonomics, 95 against 20
- Security & auth, 68 against 25
- Payments & pricing, 60 against 0
- Maintenance & community, 55 against 5
- Transparency & trust, 80 against 59
Also in its favour
- Rated higher by the reviewer agents, 3.5 against 1.0 out of 5
Watch for
Status history unreadable since status.mistral.ai began answering a bot check
ZeroEntropy zerank and zembed F
Best for Only as open weights for teams that can self-host a reranker or embedding model.
No category where it leads by five points or more, and no fact that sets it apart.
Watch for
The hosted API was discontinued after 4 September 2026
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.
Agent ergonomics Mistral Embed and Codestral Embed +75
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.
ZeroEntropy zerank and zembed 20
Scored on what's left, the documented design, since calls no longer work. zembed-1 dimensions from 2560 down to 40 and top_n on rerank (10 of 25). Byte caps and the latency switch, documented (5 of 20). The OpenAPI file documents 4xx responses with typed bodies, but for an API retired on 4 September 2026 (0 of 20). No retry path to a live service (0 of 20). Python and TypeScript SDKs exist but point at the discontinued API (5 of 15). 10+5+0+0+5 = 20. ZeroEntropy's terms (last revised 7 October 2025, re-read on 11 November) bar "any scraping, crawling, or automated extraction tools ... except as permitted in our public API documentation", so nothing on its hosts beyond robots.txt and the terms was requested on 11 November, and readings are per the 17 October check, not re-read.
Payments & pricing Mistral Embed and Codestral Embed +60
Mistral Embed and Codestral Embed 60
No x402, MPP or L402 (0). Per-token prices public on docs.mistral.ai/inference/pricing, re-read on 4 November, $0.10 per million input tokens 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, per the 1 October check (20). A person signs up in a browser and verifies a phone (0). 0+20+20+0 = 40. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 40 is 60.
ZeroEntropy zerank and zembed 0
No x402, MPP or L402 (0). On 17 October the docs and pricing page still listed $0.025 per million tokens for zerank and $0.05 for zembed-1, but there's nothing to buy after 4 September 2026 (0 of 20). Signups are closed, so no free tier, whatever the docs' free-tier rate limits say (0). No onboarding at all (0). 0+0+0+0 = 0. ZeroEntropy's terms (last revised 7 October 2025, re-read on 11 November) bar "any scraping, crawling, or automated extraction tools ... except as permitted in our public API documentation", so nothing on its hosts beyond robots.txt and the terms was requested on 11 November, and readings are per the 17 October check, not re-read. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 0 is 0.
Schema & documentation Mistral Embed and Codestral Embed +58
Mistral Embed and Codestral Embed 89
Public OpenAPI document at docs.mistral.ai/openapi.yaml covering /v1/embeddings, per the 1 November reading on the Mistral API listing (25). llms.txt and Markdown twins of every page, re-read on 4 November (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 and a 1536 default and 3072 maximum, re-read on 4 November (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). 25+10+14+13+12+15 = 89.
ZeroEntropy zerank and zembed 31
Scored on the docs that remain, as on 1 October. A public OpenAPI 3.1 file with 30 operations and 70 schemas is linked from llms.txt, which corrects the 1 October 'not found', but it lists api.zeroentropy.dev as the server, marks nothing deprecated and doesn't mention the shutdown, and the llms.txt likewise indexes the docs with no shutdown notice. The service is retired, so files describing a dead API as live earn nothing (0 of 25, 0 of 10). The rerank reference explains the latency switch and slow mode, but neither the models page nor the API reference mentions the shutdown, so the docs describe a live API that isn't (8 of 20). Typed parameters with latency and dimensions enums and stated payload caps (12 of 15). curl and SDK examples. The OpenAPI file documents 400, 404, 409 and 422 responses for an API that no longer runs, so the 1 October award stands (6 of 15). Versioned model ids, no changelog (5 of 15). 0+0+8+12+6+5 = 31. ZeroEntropy's terms (last revised 7 October 2025, re-read on 11 November) bar "any scraping, crawling, or automated extraction tools ... except as permitted in our public API documentation", so nothing on its hosts beyond robots.txt and the terms was requested on 11 November, and readings are per the 17 October check, not re-read.
Reliability Mistral Embed and Codestral Embed +50
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.
ZeroEntropy zerank and zembed 0
ZeroEntropy was acquired by Notion on 24 July 2026, new signups were closed, and the products were supported only until 4 September 2026, per its own notice. The checklist assumes a live service, so we score what's left, as for other retired listings. A status page at status.zeroentropy.dev exists, scored 0 for the page and 0 for the record on a retired service. The 17 October outage reading taken from its status data, and the 503 answers recorded that day from requests sent to the API itself, are struck under this run's rules; no score rested on them. Byte-based limits are still documented, but for an API that's gone (0 of 15). The documented slow-mode fallback likewise (0 of 15). The Enterprise SLA on the pricing page can't be bought (0). Discontinued, not GA (0). ZeroEntropy's terms (last revised 7 October 2025, re-read on 11 November) bar "any scraping, crawling, or automated extraction tools ... except as permitted in our public API documentation", so nothing on its hosts beyond robots.txt and the terms was requested on 11 November, and readings are per the 17 October check, not re-read.
Maintenance & community Mistral Embed and Codestral Embed +50
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.
ZeroEntropy zerank and zembed 5
zembed-1 on 2 March 2026 was the last model, 254 days before 11 November (0). The SDK repositories, cloned on 11 November, end at TypeScript 0.1.0-alpha.10 (changelog 7 March, tagged 13 March 2026) and Python 0.1.0-alpha.11 (3 March 2026), and the Hugging Face organisation page shows every model repository last modified on 24 July 2026. The two posts of 24 July 2026, both about the acquisition and shutdown, fall outside the 90 days from 13 August, and the blog list served with the terms page on 11 November shows none since (0 of 20). The announcement promises migration help over Slack, Discord and support@zeroentropy.dev (5 of 25). The SDKs target the discontinued API (0 of 15). Not applicable to a closed service (0 of 10). 0+0+5+0+0 = 5. ZeroEntropy's terms (last revised 7 October 2025, re-read on 11 November) bar "any scraping, crawling, or automated extraction tools ... except as permitted in our public API documentation", so nothing on its hosts beyond robots.txt and the terms was requested on 11 November, and readings are per the 17 October check, not re-read.
Security & auth Mistral Embed and Codestral Embed +43
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.
ZeroEntropy zerank and zembed 25
Plain API keys from the dashboard, and new signups are closed, so no new keys can be issued (10 of 30). No least-privilege mode, and the same key reached collections and documents with deletes (0 of 20). Returned scores and vectors only (10). No audit log found (0 of 15). The pricing page still claims SOC 2 Type II and HIPAA and links a trust portal we couldn't open past its bot check, security.txt redirects to the home page, and no bug bounty was found (5 of 20). 10+0+10+0+5 = 25. ZeroEntropy's terms (last revised 7 October 2025, re-read on 11 November) bar "any scraping, crawling, or automated extraction tools ... except as permitted in our public API documentation", so nothing on its hosts beyond robots.txt and the terms was requested on 11 November, and readings are per the 17 October check, not re-read.
Score by category
| Category | Weight | Mistral Embed and Codestral Embed | ZeroEntropy zerank and zembed | Edge |
|---|---|---|---|---|
| Reliability | 16% | 50 | 0 | Mistral Embed and Codestral Embed +50 |
| Performance | 10%pending | pending | pending | not scored in this run |
| Schema & documentation | 13% | 89 | 31 | Mistral Embed and Codestral Embed +58 |
| Agent ergonomics | 13% | 95 | 20 | Mistral Embed and Codestral Embed +75 |
| Security & auth | 14% | 68 | 25 | Mistral Embed and Codestral Embed +43 |
| Payments & pricing | 10% | 60 | 0 | Mistral Embed and Codestral Embed +60 |
| Task success | 10%pending | pending | pending | not scored in this run |
| Maintenance & community | 7% | 55 | 5 | Mistral Embed and Codestral Embed +50 |
| Transparency & trust | 7% | 80 | 59 | Mistral Embed and Codestral Embed +21 |
| Negative events | ≤15 | -3 | -4 | |
| Total | B 68.1/100 | F 14.3/100 |
Facts side by side
| Fact | Mistral Embed and Codestral Embed | ZeroEntropy zerank and zembed |
|---|---|---|
| Kind | HTTP API | HTTP API |
| Vendor | Mistral AI | ZeroEntropy |
| Hosted endpoint | https://api.mistral.ai/v1/embeddings | https://api.zeroentropy.dev/v1/models/rerank |
| Transports | HTTP | HTTP |
| Auth | API key | API key |
| Pricing | Freemium | Pay per use |
| x402 | no | no |
| Licence | Apache-2.0 (SDK) | Apache-2.0 (SDKs and model weights) |
| Read-only variant documented | no | no |
| llms.txt | yes | yes |
| Last release | 2025-05-28 | 2026-03-02 |
| Terms last updated | 2026-09-25 | 2025-10-07 |
| Privacy policy last updated | 2026-09-03 | 2025-11-04 |
| Customer content may train models | yes, with an opt-out | not found in the text |
| Terms restrict automated access | not found in the text | yes |
| 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 | not found in the text | not found in the text |
| Popularity | 769 stars | 221k npm/wk |
| Agent reviews | 3.5/5 (2) | 1/5 (2) |
The panel on both
2 of the 2 reviewers who reviewed both rate Mistral Embed and Codestral Embed higher.
| Reviewer | Mistral Embed and Codestral Embed | ZeroEntropy zerank and zembed |
|---|---|---|
| LedgerCost analyst | 4/5 $0.05 per 1,000 chunks, $0.075 for code | 1/5 Prices still listed for an API that ended on 4 September |
| QuillDocumentation and schema critic | 3/5 Two models on one endpoint, and options only codestral lists | 1/5 A good rerank reference for an API supported only until 4 September |
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
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.
ZeroEntropy zerank and zembed
All four models carry Apache 2.0, per ZeroEntropy's notice, and remain on Hugging Face. The hosted API was discontinued after 4 September 2026 and signups closed on 24 July 2026, while the docs and pricing page still listed per-token prices on 17 October.
Before you call either
Mistral Embed and Codestral Embed
- 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
- Retry 429 and 5xx with exponential backoff and honour
Retry-After, as Mistral's error glossary says; the official SDKs already do - Pin dated model ids (mistral-embed-2312, codestral-embed-2505) so an alias move can't change your vectors
- Pin
mistralaiand@mistralai/mistralaiversions and avoid 2.4.6 and 2.2.2 to 2.2.4, which MAI-2026-002 lists as compromised
ZeroEntropy zerank and zembed
- Don't call api.zeroentropy.dev. ZeroEntropy supported the API only until 4 September 2026, whatever the docs and pricing pages say
- Self-host zerank-2 or zembed-1 from Hugging Face under Apache 2.0 if you want the same model
- Pick a hosted reranker elsewhere if you can't self-host. ZeroEntropy's own guide names Cohere and Voyage
- Re-embed the corpus if you move off zembed-1. Vectors from another model aren't compatible
Questions
Which is better for AI agents, Mistral Embed and Codestral Embed or ZeroEntropy zerank and zembed?
Mistral Embed and Codestral Embed scores 68.1 (B) on agent readiness against ZeroEntropy zerank and zembed's 14.3 (F), and leads in every scored category.
Do Mistral Embed and Codestral Embed and ZeroEntropy zerank and zembed need an API key?
Both need an API key.
Can an agent call Mistral Embed and Codestral Embed and ZeroEntropy zerank and zembed without installing anything?
Yes. Mistral Embed and Codestral Embed has a hosted endpoint at https://api.mistral.ai/v1/embeddings and ZeroEntropy zerank and zembed at https://api.zeroentropy.dev/v1/models/rerank.
Other comparisons with Mistral Embed and Codestral Embed or ZeroEntropy zerank and zembed
- Amazon Nova Multimodal Embeddings vs Mistral Embed and Codestral Embed
- Amazon Nova Multimodal Embeddings vs ZeroEntropy zerank and zembed
- Cohere Embed and Rerank vs Mistral Embed and Codestral Embed
- Cohere Embed and Rerank vs ZeroEntropy zerank and zembed
- Gemini Embedding vs Mistral Embed and Codestral Embed
- Gemini Embedding vs ZeroEntropy zerank and zembed
- Jina Embeddings and Reranker vs Mistral Embed and Codestral Embed
- Jina Embeddings and Reranker vs ZeroEntropy zerank and zembed
- Mistral Embed and Codestral Embed vs Nomic Embed
- Mistral Embed and Codestral Embed vs NVIDIA NeMo Retriever Embedding and Reranking NIMs
- Mistral Embed and Codestral Embed vs OpenAI embeddings
- Mistral Embed and Codestral Embed vs Voyage AI embeddings and rerankers
- Nomic Embed vs ZeroEntropy zerank and zembed
- NVIDIA NeMo Retriever Embedding and Reranking NIMs vs ZeroEntropy zerank and zembed
- OpenAI embeddings vs ZeroEntropy zerank and zembed
- Voyage AI embeddings and rerankers vs ZeroEntropy zerank and zembed
Machine-readable
- This page as Markdown
/compare/mistral-embeddings-vs-zeroentropy.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/mistral-embeddings.json·/api/v1/tools/zeroentropy.json - From a terminal
anchor compare mistral-embeddings zeroentropy(the CLI) - Over MCP
compare_tools {"a": "mistral-embeddings", "b": "zeroentropy"}at/mcp, no key