Skip to content
aicoolies logo
DeepSeek logo

DeepSeek Pricing

freemiumopen sourcelast verified October 18, 2026official site ↗
PlanPriceNotes
DeepSeek Web Chatofficial source ↗$0checked 2026-08-26Free web chat interface access to DeepSeek models (subject to standard web concurrency rate limits)
DeepSeek APIofficial source ↗Usage-basedchecked 2026-11-03Per 1M tokens, off-peak/peak: Flash (DeepSeek-V4.1-Flash-1017) input cache hit $0.003/$0.006, cache miss $0.15/$0.30, output $0.60/$1.20; V4-Pro input cache hit $0.022/$0.044, cache miss $0.66/$1.32, output $1.98/$3.96. Off-peak rates are half of peak; peak hours 01:00-04:00 and 06:00-10:00 UTC Mon-Fri (excluding Chinese public holidays). Automatic context caching.
Search & Fetch APIofficial source ↗Usage-basedchecked 2026-11-03Search $0.75 and Fetch $0.75 per 1,000 successful requests. Server-side Search/Fetch calls made through DeepSeek Harness, Claude Code or Codex were free until November 1, 2026 at 00:00 Beijing time (October 31, 16:00 UTC); since then they are billed at the same per-request rates.

structured from cited official pricing sources: “DeepSeek provides open-weights models and high-efficiency direct API endpoints. The web chat interface is free, while the official API is usage-based per million tokens with automatic context caching and off-peak rates at half the peak price ($0.15-$1.32 cache-miss input and $0.60-$3.96 output per 1M tokens across Flash and Pro). The standalone Search and Fetch APIs cost $0.75 per 1,000 successful requests each.”

Self-hosting cost

DeepSeek is open source. Self-hosting does not add a software license fee, but you still pay for infrastructure, operations, and any external model or API usage. Any hosted, managed, or API plans remain separately priced.

Free & cheaper Model Providers

same category, free plan or open-source license

ClaudeFree tier with core Claude access. Paid subscription plans include Claude Pro at $20/month ($17/month billed annually), Claude Max at $100/month (5x Pro usage), Claude Ultra at $200/month (formerly Max 20x; 10x Pro usage every week, no 5-hour limit for verified accounts), and Claude Team (2–150 seats) with Standard seats at $25/seat/month ($20 billed annually) and Premium seats at $125/seat/month ($100 billed annually, 5x Standard usage). Enterprise is $20/seat/month billed annually plus usage at API rates, with SSO, SCIM provisioning and audit logs. Since October 7, 2026, Max and Team plans also include monthly Claude Platform API credits: $100 on Max, $200 on Ultra, and on Team $20 per Standard seat and $100 per Premium seat, pooled and capped at $500. The credits cover the Claude API, Managed Agents and the Agent SDK, not interactive Claude Code, and expire each billing cycle.CerebrasCerebras Inference has a self-serve Developer tier for development and evaluation, billed pay-as-you-go with a one-time $5 promotional credit (requires a valid payment method; expires after 30 days). Developer models are GPT OSS 120B at $0.35/1M input and $0.75/1M output tokens and Qwen 3.8 27B at $0.99/1M input and $1.49/1M output. Enterprise is quote-based for production workloads, with all supported models, higher priority and rate limits, enhanced latency options, model-weight customization and fine-tuning services.ChatGPTChatGPT plans (vendor_documented, as-of 2026-09-14): Free; Go (~$8/mo US); Plus $20/mo; Pro from $100/mo (5×) and $200/mo (20×—new $200 sign-ups paused as of 2026-09-10); Business Standard $25/user/mo or $20/user/mo annual (Premium seats $125/$100 annual); Enterprise custom. Product URL: chatgpt.com. Former “Team” plan is now Business. Confirm geo-localized Go/Plus/Pro card amounts on https://chatgpt.com/pricing/. OpenAI says that starting in December 2026, the usage allowance of Plus, Pro and Business plans will be shared between ChatGPT and Codex and measured in tokens, with more advanced models and complex tasks using more of it; when it runs out you can wait for the reset, upgrade or buy credits. Instant chat does not count against the shared allowance.GroqGroqCloud has a Free plan with rate-limited API access and a pay-as-you-go Developer plan with higher rate limits plus Batch and Flex processing. Self-serve prices include GPT-OSS 20B at $0.075/1M input and $0.30/1M output tokens, GPT-OSS 120B at $0.15/$0.60 and Qwen3.8-27B (preview) at $0.80/$4.00; Whisper speech-to-text starts at $0.04 per audio hour. Cached input tokens are discounted 50% and Batch API requests cost 50% less. Llama 3.1 8B, Llama 3.3 70B and MiniMax M2.7 are offered through enterprise sales, and dedicated capacity is sold as GroqMetal, GroqCore and GroqAssured.

FAQ

How much does DeepSeek cost?

DeepSeek provides open-weights models and high-efficiency direct API endpoints. The web chat interface is free, while the official API is usage-based per million tokens with automatic context caching and off-peak rates at half the peak price ($0.15-$1.32 cache-miss input and $0.60-$3.96 output per 1M tokens across Flash and Pro). The standalone Search and Fetch APIs cost $0.75 per 1,000 successful requests each. Paid pricing starts at $0.75 per 1,000 successful requests.

Does DeepSeek have a free plan?

Yes — DeepSeek has a free plan alongside its paid tiers.

Can you self-host DeepSeek for free?

DeepSeek is open source. Self-hosting does not add a software license fee, but you still pay for infrastructure, operations, and any external model or API usage. Any hosted, managed, or API plans remain separately priced.

What are free or cheaper DeepSeek alternatives?

Claude, Cerebras, ChatGPT are in the same category with a free plan or open-source license.

pricing is editorial catalog data (last verified October 18, 2026) — vendors change plans; confirm on the official site before you buy. Reuse this data under CC-BY 4.0 via /data.