DeepSeek through Squiid

Low-cost chat and reasoning models with an OpenAI-compatible API. DeepSeek is on the Squiid roadmap.

Roadmap Instant Instant access. A usage-metered API that works the moment you add it to your dashboard.

What DeepSeek is

DeepSeek publishes chat and reasoning model families and serves them from its own platform with an API that follows the OpenAI request format. Two things drive its adoption: price per million tokens that is a fraction of frontier providers, and a context caching system that charges less for prompt prefixes it has already seen, which matters when every request resends the same long system prompt.

In an agent-built product DeepSeek tends to be the volume tier. Tagging thousands of records, generating first-draft descriptions for a catalog, running an evaluation suite overnight, or backfilling summaries across a database are all jobs where a ten times price difference decides whether the feature ships. Its reasoning model is also used as a cheaper checker in front of a more expensive model. The reasoning models expose their chain of thought separately from the final answer, which makes them usable as a grader in evaluation harnesses that an agent writes alongside the feature.

Why connect it through Squiid

Because those are exactly the workloads that run in loops, a hard spending boundary is useful. Squiid meters DeepSeek calls against prepaid credits and stops authorizing them when the balance hits zero, so a runaway batch job costs you what you funded and nothing more. Requests go to https://api.squiid.io/v1/deepseek/... with Authorization: Bearer $SQUIID_API_KEY, any OpenAI-compatible client works by pointing its base URL at the gateway (gateway convention), and the spend lands on the same monthly invoice as the rest of your stack. Spend per day and per model is visible while the job is still running, not at the end of the month.

  • One account, one key, one bill, one dashboard. DeepSeek sits beside every other service your project uses, on the same invoice and the same credit balance.
  • Key custody. Squiid holds the DeepSeek credentials. Your code carries SQUIID_API_KEY, so DEEPSEEK_API_KEY, DEEPSEEK_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 DeepSeek 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

  • DeepSeek chat and reasoning models
  • OpenAI-compatible calls through the gateway
  • Cached prefix pricing passed through as billed
  • Hard stop at zero credits for batch jobs
  • DeepSeek list pricing with no markup

How to connect DeepSeek

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 DEEPSEEK_API_KEY, DEEPSEEK_BASE_URL
SQUIID_API_KEY=sq_live_xxxxxxxxxxxxxxxxxxxxxxxx

# every request: https://api.squiid.io/v1/deepseek/...
# header:       Authorization: Bearer $SQUIID_API_KEY
01

Add it in your dashboard

Open the Squiid catalog, add DeepSeek and fund credits. Access is instant and metered per request.

02

Put the one key in your env

Add SQUIID_API_KEY to .env and delete DEEPSEEK_API_KEY, DEEPSEEK_BASE_URL. Nothing else from DeepSeek needs to exist in your project.

03

Point your agent at it

Claude Code, Cursor and Codex read the same variable and call https://api.squiid.io/v1/deepseek/. Agents usually point the existing OpenAI client at DeepSeek for bulk or background jobs while keeping a frontier model on the interactive path.

Pricing through Squiid

DeepSeek charges per input and output token with a lower rate for cache hits on repeated prompt prefixes, at the provider's listed price. Squiid passes that through at par: one dollar of prepaid credit is one dollar of DeepSeek usage, with no markup and no per-request margin.

DeepSeek 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 DeepSeek 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: DeepSeek is usage-metered, so leaving means creating your own DeepSeek key and pointing the client at it. Squiid never owns your code, your schema or your domain, and the gateway speaks DeepSeek'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

DeepSeek 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:

Vibe coding a feature on DeepSeek 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

Questions people ask

Can I run a big batch job on DeepSeek without risking a huge bill?

Yes. Squiid meters every call against prepaid credits and refuses new calls at zero, so the job stops instead of running up an unfunded invoice.

Does the OpenAI SDK work with DeepSeek through Squiid?

Yes. DeepSeek keeps the OpenAI request shape, so pointing the client's base URL at the DeepSeek path on the Squiid gateway is enough.

Does Squiid mark up DeepSeek's low prices?

No. A dollar of credit buys a dollar of DeepSeek usage. The Squiid plan fee and top-up percentage are separate and visible.

Is context caching supported?

Yes. Cache behaviour is DeepSeek's own, and whatever rate DeepSeek bills for a cache hit is what is metered against your credits.

Where does my data go?

Requests are forwarded to DeepSeek's API and are subject to DeepSeek's own data terms. Squiid does not add training rights or retention of its own.

Connect DeepSeek with one key.

DeepSeek serves its chat and reasoning models over an OpenAI-compatible API at prices well below frontier providers, with context caching on repeated prompts. Metered per token through Squiid.