Cohere through Squiid
Enterprise models with strong embeddings, rerank and RAG tooling. Cohere is on the Squiid roadmap.
What Cohere is
Cohere sells language models aimed at enterprise retrieval workloads. The Command family handles generation and tool use, Embed produces multilingual embeddings for vector search, and Rerank takes a query and a candidate list and reorders it by relevance. That last one is the piece most teams end up buying: a rerank pass on top of an ordinary vector search usually improves answer quality more than swapping the generation model does.
In an AI-coded app Cohere typically sits in the middle of a retrieval pipeline. Documents are chunked and embedded with Embed, stored in a vector database, retrieved by similarity, reranked by Cohere, and only then passed to whichever model writes the answer. Because rerank is called once per user query with a list of candidates, it is cheap relative to generation and easy to add late. Cohere's models are also offered through private deployments for regulated customers, which is why it shows up in enterprise stacks more often than in weekend projects.
Why connect it through Squiid
Cohere is on the Squiid roadmap. When it lands, requests will go to https://api.squiid.io/v1/cohere/... with Authorization: Bearer $SQUIID_API_KEY, the official Cohere SDKs will work by pointing their base URL at the gateway (gateway convention), and embedding, rerank and generation calls will all meter against the same prepaid credits at Cohere's list price. Squiid will hold the upstream key, so a rotation in the dashboard cuts off every agent and deployment at once. Embed, rerank and generation will be itemised separately in the usage view so you can see which stage of the pipeline costs what.
- One account, one key, one bill, one dashboard. Cohere sits beside every other service your project uses, on the same invoice and the same credit balance.
- Key custody. Squiid holds the Cohere credentials. Your code carries
SQUIID_API_KEY, so COHERE_API_KEY, CO_API_KEY, COHERE_BASE_URL never land in a repository, a preview deployment or an agent’s context. - Metering you can see. Every call is counted per service and per day, so you know what Cohere costs before the month ends.
- Pause on zero. When prepaid credits run out the gateway stops authorizing usage instead of running up a bill you did not fund.
- One rotation. Rotating the Squiid key cuts off every agent, worker and deployment at once, with no provider key to hunt down.
What you get
- Command generation models through the gateway
- Embed multilingual embeddings on the same key
- Rerank for retrieval quality without a second account
- Metering per endpoint against prepaid credits
- Cohere list pricing, no markup
How to connect Cohere
One variable replaces the provider credentials this service would normally need. Requests go to the gateway and the official SDKs work by pointing their base URL at it (gateway convention).
# .env
# replaces COHERE_API_KEY, CO_API_KEY, COHERE_BASE_URL
SQUIID_API_KEY=sq_live_xxxxxxxxxxxxxxxxxxxxxxxx
# every request: https://api.squiid.io/v1/cohere/...
# header: Authorization: Bearer $SQUIID_API_KEY
Add it in your dashboard
Open the Squiid catalog, add Cohere and fund credits. Access is instant and metered per request.
Put the one key in your env
Add SQUIID_API_KEY to .env and delete COHERE_API_KEY, CO_API_KEY, COHERE_BASE_URL. Nothing else from Cohere needs to exist in your project.
Point your agent at it
Claude Code, Cursor and Codex read the same variable and call https://api.squiid.io/v1/cohere/. Agents wire Cohere Rerank between the vector search call and the prompt, usually as a five line addition to an existing retrieval function.
Pricing through Squiid
Cohere charges per token for generation and embeddings and per search unit for rerank, at the provider's listed price. Squiid passes that through at par: one dollar of prepaid credit is one dollar of Cohere usage, with no markup and no per-request margin.
Cohere has no subscription attached on Squiid today. If a plan is required for a feature you need, it passes through at the provider’s listed price with no Squiid fee.
Squiid adds no markup to provider prices. There is a free plan at $0 a month with a 15% fee when you top up credits, Solo at $19.97 a month with a top-up fee that starts at 8% and falls to 3.5% as your spend grows, Team from $29.97 a seat a month with the same ladder one point lower, and Custom with a rate of 3% or less. Credits are prepaid, so Cohere usage stops when the balance reaches zero rather than producing an invoice you did not fund. Auto top-up with a cap keeps production running without handing anyone an open-ended card.
If you leave
Nothing to move: Cohere is usage-metered, so leaving means creating your own Cohere key and pointing the client at it. Squiid never owns your code, your schema or your domain, and the gateway speaks Cohere's own API, so the client change is a base URL. The handover page lists what happens for every provider in the catalogue.
Works with every coding agent
Cohere through Squiid is just an HTTP call with a bearer token, so every agentic coding tool can use it with the single key already in your environment:
- Claude Code
- Cursor
- ChatGPT
- Codex
- Grok
- Gemini CLI
- GitHub Copilot
- Windsurf
- Perplexity
- Devin
- Replit
- v0
- Lovable
- Cline
Vibe coding a feature on Cohere usually means asking for it in plain language and letting the agent write the integration. Because the key is Squiid’s rather than the provider’s, it can do that without ever holding a credential that could be leaked, logged or committed.
Related services
Pinecone
Vector and searchManaged vector database for retrieval augmented generation at scale
Next upProvisionedQdrant Cloud
Vector and searchOpen-source vector search with rich payload filtering and quantization
RoadmapProvisionedOpenAI
LLM APIsGPT models, embeddings, images and realtime voice over one API
LiveInstant WEWeaviate
Vector and searchVector database with built-in vectorization modules and hybrid search
RoadmapProvisionedMistral
LLM APIsEuropean open-weight and frontier models, plus OCR and embeddings
Next upInstantQuestions people ask
When will Cohere be available on Squiid?
It is on the roadmap rather than rolling out. Add it to your dashboard watchlist and you will be notified when provisioning opens.
Why would I add Cohere if I already have OpenAI embeddings?
Mainly for Rerank, which reorders retrieved chunks by relevance and usually improves answer quality more cheaply than changing the generation model.
Will Rerank bill separately from generation?
Yes. Cohere prices rerank per search unit and generation per token. Both will be metered against the same Squiid credits at list price.
Can Claude Code build a RAG pipeline with Cohere through Squiid?
Once Cohere is live, yes. The agent writes embed, search and rerank calls that all carry the single SQUIID_API_KEY.
Does Squiid support Cohere's multilingual embeddings?
Yes, the plan is to pass the full Embed model list through the gateway with no model-level restrictions.
Connect Cohere with one key.
Cohere builds the Command generation models plus Embed and Rerank, which are widely used to improve retrieval quality in RAG pipelines. Planned for Squiid with usage-based metering and no markup.