How do I use DeepSeek in Cline?
In Cline, set API Provider to OpenAI Compatible, Base URL to https://api.stdcmpt.com/v1, paste a Standard Compute key and enter deepseek/deepseek-v4.1-flash as the Model ID: Cline then runs on DeepSeek V4.1 Flash on a plan from $19 a month, and new requests get HTTP 402 when the plan's compute budget reaches 100%.
# Act mode: DeepSeek V4.1 Flash
API Provider OpenAI Compatible
Base URL https://api.stdcmpt.com/v1
OpenAI Compatible API Key YOUR_API_KEY
Model ID deepseek/deepseek-v4.1-flash
# MODEL CONFIGURATION
Supports Images on
Context Window Size 400000
Max Output Tokens 32000
Reasoning Effort medium
Supports Images starts ticked for every OpenAI Compatible model; untick it for V4 Pro, or a screenshot in Plan mode fails with HTTP 400. The CLI reads the key from the variable you pass with -k.
Plan on V4 Pro, act on Flash
Tick "Use different models for Plan and Act modes" and Cline keeps a separate model, and a separate Reasoning Effort, per mode. Plan mode on deepseek/deepseek-v4-pro reasons through the approach; Act mode on V4.1 Flash does the many edit and command turns. At DeepSeek's peak price V4 Pro costs about 5 times as much per day.
Three settings Cline gets wrong by default
| Setting | Cline's default | Set it to |
|---|---|---|
| Supports Images | Ticked for every model | Ticked for V4.1 Flash, unticked for V4 Pro, which answers an image with HTTP 400 |
| Context Window Size | 128,000 | 400,000; Cline compacts or truncates as the session nears it, and the endpoint takes about 850,000 |
| Reasoning Effort | none | medium or high. Cline's value overrides the Models page setting for every request |
From Cline's OpenAI Compatible settings source (OpenAICompatible.tsx, api.ts), checked October 14, 2026. Unset prices make Cline's cost display show $0; billing is unaffected.
What a day of DeepSeek costs in Cline
| Per day | V4.1 Flash, peak | Off-peak | V4 Pro, peak |
|---|---|---|---|
| Typical day | $0.42 | $0.21 | $1.94 |
| Busy day | $4.31 | $2.16 | $20 |
DeepSeek list prices per million tokens: V4.1 Flash $0.30 input, $0.006 cache hit, $1.20 output at peak, half off-peak; V4 Pro $1.32, $0.044, $3.96 at peak. Source. A typical day is 8.5M tokens and a busy one 86.7M, 87.2% of input as cache reads: the median and 90th-percentile coding-agent day in our Agent Burn Index, measured on OpenCode, Cline, Kilo and similar agents, September 13 to October 12, 2026. Priced at the list rates above; your own days will differ.
The catch
ClinePass, Cline's own $9.99-a-month plan, includes DeepSeek V4.1 Flash and V4 Pro (cline-pass/deepseek-v4.1-flash); Cline describes it as 2 to 5 times standard API usage but does not publish the size of its 5-hour, weekly and monthly windows. Cline's built-in DeepSeek provider takes a DeepSeek key (model deepseek-flash) at the prices above. For DeepSeek alone, both cost less than a plan.
Where Standard Compute fits
A published compute budget, and Plan mode can move from V4 Pro to Claude Sonnet 5.5 by changing one Model ID, on the same key and the same Base URL.
FAQ
Can I use deepseek-flash as the Model ID?
Not on this endpoint: deepseek-flash is the name on DeepSeek's own API and in Cline's built-in DeepSeek provider. Here the ID is deepseek/deepseek-v4.1-flash, or the short form deepseek-v4.1-flash.
Why does Cline show $0.00 for my DeepSeek requests?
Cline's OpenAI Compatible provider prices unknown models at $0 unless you fill in Input and Output Price under Model Configuration. It is display only; usage counts against your plan's compute budget.
Does the Cline CLI work with this?
Yes: cline auth -p openai-compatible -k with your key, -m deepseek/deepseek-v4.1-flash and -b https://api.stdcmpt.com/v1. The -b flag only works for OpenAI and OpenAI-compatible providers.