#20 in Chat AssistantsUpdated 2026-08

DeepSeek

by DeepSeek · Chat and API access to the DeepSeek V4 models, billed per token

30.9 — BenchRank score out of 100

Homepage
Pricing page

Visit website See full Chat Assistants ranking

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

Pricing

Published plans from DeepSeek’s own pricing page, in USD. Usage charges and add-ons may apply on top.

DeepSeek pricing tiers, monthly rates in USD
Plan	Monthly	Includes
deepseek-v4-flash	—per 1M tokens	$0.14 per 1M input tokens (cache miss), $0.0028 (cache hit) $0.28 per 1M output tokens Concurrency limit 2500; supports the Responses API
deepseek-v4-pro	—per 1M tokens	$0.435 per 1M input tokens (cache miss), $0.003625 (cache hit) $0.87 per 1M output tokens Concurrency limit 500; no Responses API support yet

How this score is made up

Each dimension is scored out of 100 and combined into the headline score using fixed weights.

MCP support: 0 out of 100
API quality: 0 out of 100
Documentation: 0 out of 100
Agent friendliness: 50 out of 100
Pricing transparency: 65 out of 100
Customer sentiment: 80 out of 100
Changelog: 40 out of 100
Marketing site structure: 65 out of 100
Operational trust: 25 out of 100

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: 60 out of 100
Maintenance: 85 out of 100

This doesn’t look right — report a problem with DeepSeek’s score

Alternatives in Chat Assistants

Ranked 1
77.1 — BenchRank score out of 100
Browser 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 2
76.6 — BenchRank score out of 100
LobeHub
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 3
68.8 — BenchRank score out of 100
Open WebUI
Open WebUI · Self-hosted interface for running local and cloud AI models
Best for: Teams that want to self-host one interface over local and cloud models on their own infrastructure.