Setup guide - From DeepSeek and Roo Code docs
Use DeepSeek in Roo Code
Roo Code's built-in DeepSeek provider still ships only the discontinued deepseek-chat / deepseek-reasoner ids. To use current models, use Roo Code's documented OpenAI Compatible provider with DeepSeek's documented OpenAI-format base URL and the model id deepseek-flash or deepseek-v4-pro.
Last verified 5 steps6 official sources
Steps
- Step 1.
Create an API key at https://platform.deepseek.com/api_keys.
- Step 2.
Open the Roo Code settings panel and set API Provider to "OpenAI Compatible".
- Step 3.
Set Base URL to DeepSeek's OpenAI-format base URL.
Snippethttps://api.deepseek.com - Step 4.
Enter your DeepSeek API key in API Key.
- Step 5.
Set Model to a current DeepSeek model id. Optionally, under Model Configuration set Context Window and Max Output Tokens (DeepSeek documents 1M context and 384K max output), and turn on Image Support only for deepseek-flash.
Snippetdeepseek-flash
Caveats
- This combines Roo Code's documented OpenAI Compatible mechanism with DeepSeek's documented base URL and model ids. Neither vendor documents this exact pairing.
- Roo Code's DeepSeek provider page and its source (packages/types/src/providers/deepseek.ts on main) list only deepseek-chat (default) and deepseek-reasoner. DeepSeek's change log announced those names would be discontinued on 2026-07-24.
- Roo Code only uses native tool calling. In DeepSeek thinking mode (on by default), requests that include tools must pass reasoning_content back on every later request, or the API returns a 400 error. Roo's docs do not say whether the OpenAI Compatible provider does this. If you see 400s, this is the likely cause.
More setup guides
Same tool, other providers
Related reading
- DeepSeek V4.1-Flash-1017: Benchmarks vs Opus 5, Pricing, Switch?DeepSeek V4.1-Flash-1017 jumps 11 points on Terminal-Bench 4.0 at the same deepseek-flash price. Benchmarks vs Opus 5, K3, GLM-5.3 and when to switch.
- DeepSeek 4.1 Flash as a Daily Coding Model: Cache, Peak Hours and LimitsDeepSeek 4.1 Flash as a daily coding model: what a session costs, how peak hours double the bill, why caching matters, and where the subsidy argument holds.
- Where to Run DeepSeek V4.1 Flash Free and CheapDeepSeek V4.1 Flash replaced V4 Flash. Official API: $0.15 input, $0.60 output off-peak. OpenRouter hosts start near $0.05. MIT weights are on Hugging Face.
- DeepSeek V4 Flash Vision Exp: Experimental Vision, Limits, and How to Run It in OpenCodeDeepSeek V4 Flash Vision Exp guide: supported image formats, three input methods, the 384-token per-image cap, and how to run it in OpenCode today.
- We Read DeepSeek Harness: What 453K Lines of Agent Runtime Actually SayDeepSeek Harness (dsh) code read: a 453K-line plugin runtime on a vendored Cordis fork, three patterns worth stealing, and the gaps the launch does not mention.
- DeepSeek V4 Flash Is 90% Off Through Novita on Vercel AI Gateway: The Cost MathDeepSeek V4 Flash via Novita on Vercel AI Gateway was 90% off for Pro customers through August 11. The before/after cost math and the provider-pinning setup.
FAQ
How do I use DeepSeek in Roo Code?
Roo Code's built-in DeepSeek provider still ships only the discontinued deepseek-chat / deepseek-reasoner ids. To use current models, use Roo Code's documented OpenAI Compatible provider with DeepSeek's documented OpenAI-format base URL and the model id deepseek-flash or deepseek-v4-pro.
Which DeepSeek model ids work in Roo Code?
The official docs read for this guide list deepseek-flash and deepseek-v4-pro. Requests go to https://api.deepseek.com.
Is this an official DeepSeek and Roo Code setup?
The steps draw on both DeepSeek's and Roo Code's docs. It was last checked against 6 official pages on October 17, 2026.
Sources
Last verified against 6 official pages.
- docs.roocode.comhttps://docs.roocode.com/providers/openai-compatible
- docs.roocode.comhttps://docs.roocode.com/providers/deepseek
- raw.githubusercontent.comhttps://raw.githubusercontent.com/RooCodeInc/Roo-Code/main/packages/types/src/providers/deepseek.ts
- api-docs.deepseek.comhttps://api-docs.deepseek.com/
- api-docs.deepseek.comhttps://api-docs.deepseek.com/guides/thinking_mode
- api-docs.deepseek.comhttps://api-docs.deepseek.com/updates
Newsletter
New models, set up the day they ship
One email per week. New model releases, the coding tools that support them and how to wire them up.