DeepSeek
by DeepSeek · Chat and API access to the DeepSeek V4 models, billed per token
BenchRank score
Screenshots of DeepSeek
Homepage
Overview
DeepSeek offers chat models through a web app and a token-billed API. Two models are published, deepseek-v4-flash and deepseek-v4-pro, both with a 1M-token context, 384K maximum output, and thinking and non-thinking modes. The API is reachable in OpenAI or Anthropic format and supports JSON output, tool calls and context caching.
- Best for
- Developers wanting a token-billed LLM API with OpenAI- and Anthropic-compatible endpoints
- Pricing
- No subscription is shown; the API bills per token, with deepseek-v4-flash at $0.14 per 1M input tokens (cache miss) and $0.28 per 1M output, and deepseek-v4-pro at $0.435 and $0.87, while the chat app is described as free.
Strengths and trade-offs
Strengths
- 1M-token context, up to 384K output tokens on both models
- Callable in OpenAI format or Anthropic format
- Cache-hit input tokens billed far below cache-miss rates
- Free access to the DeepSeek chat model via web and app
Trade-offs
- Peak-hour pricing will double all billing items (09:00-12:00, 14:00-18:00 UTC+8)
- Responses API does not support deepseek-v4-pro until early August 2026
- deepseek-v4-pro is capped at 500 concurrent requests versus 2500 for Flash
- FIM and prefix completion are beta; FIM is non-thinking mode only
How DeepSeek markets itself
A structured read of the promise, proof and page design on DeepSeek’s captured homepage.
Homepage capture
“Into the unknown”
- Angle: Curiosity / intrigue
- Hero: Abstract graphic
Pricing
Published plans and prices from DeepSeek’s own pricing page.
How this score is made up
Each dimension is scored out of 100 and combined into the headline score using fixed weights.
MCP support
Whether an agent can drive the product through the Model Context Protocol, and how much setup that takes.
API quality
Public API surface: machine-readable spec, official SDKs, documented auth, errors, rate limits and versioning.
Documentation
Publicly reachable docs — coverage, freshness, code samples and machine readability.
Agent friendliness
How readable the site is to an automated client: llms.txt, structured data, server-rendered content, crawler access.
Pricing transparency
Whether real prices are published, self-serve signup exists, and usage costs are knowable without a sales call.
Customer sentiment
Third-party ratings, shrunk toward the catalogue mean by review volume and corroborated across independent sources.
Changelog
A public, dated record of what shipped and when — the clearest signal that a product is still alive.
Marketing site structure
Whether the site answers a buyer's questions: clear positioning, the pages that matter, and accessibility.
Page speed
How fast the site loads for real visitors: Chrome UX Report 75th-percentile LCP, INP and CLS, with a Lighthouse mobile run standing in where a site has too little traffic for field data.
Operational trust
Status page and incident history, security disclosure, compliance and data-processing documentation.
Measured, but not part of the score
Useful to know, but not a mark for or against the product — so these do not affect the ranking.
Openness
Source availability, self-hosting, data export and open standards. Scored and shown, but not part of the composite — paid SaaS is not worse for being paid SaaS.
Maintenance
Release cadence and repository activity. Scored and shown, but not part of the composite — it is only measurable for open repositories.
This doesn’t look right — report a problem with DeepSeek’s score
Where this comes from
The DeepSeek pages BenchRank reads when it scores the product — its documentation, release notes, status and security pages, and its repository where there is one.
Alternatives in Chat Assistants
Ranked 1
77.8 — BenchRank score out of 100LobeHub
LobeHub · Hosted multi-agent platform with a skills and MCP marketplace
Best for: Individuals and small teams wanting to run several AI agents on a hosted, credit-metered platform
Ranked 2
77.3 — BenchRank score out of 100Browser Use
Browser Use · Open-source browser automation with hosted stealth browsers and agents
Best for: Teams building web automation that need hosted stealth browsers and agent APIs, plus an open-source library
Ranked 3
71.3 — BenchRank score out of 100OpenClaw
OpenClaw · Open-source AI assistant that runs on your own machine
Best for: People who want an AI assistant they run themselves and drive from WhatsApp, Telegram or Discord.
Is this your product?
Claim DeepSeek to manage its profile. Claiming lets you suggest edits to the descriptive fields — it never changes scores or rankings.