DeepSeek Flash Lite Xploded in Continue

DeepSeek Flash Lite Xploded in Continue

Run deepseek-v4.1-flash-lite-1017-xploded in Continue through MagmaRouter: one key, a hard spend cap, and a model that answers instead of refusing.

Model iddeepseek-v4.1-flash-lite-1017-xploded
Context window512,000 tokens (512K)
Tool callingSupported
Price per 1M tokens$0.36 input, $0.030 on a cache hit, $1.44 output
Intelligence39 on the Artificial Analysis Intelligence Index v4.3.2, for the base model
Web searchAdd :online to the id

Set it up

  1. Add a model block to ~/.continue/config.yaml with model: deepseek-v4.1-flash-lite-1017-xploded and the MagmaRouter apiBase.
  2. Put MAGMAROUTER_API_KEY=sk-mr-... in ~/.continue/.env; the editor extensions cannot read your shell.
  3. Keep capabilities: [tool_use], or Continue will not let it act in agent mode.
name: MagmaRouter
version: 0.0.1
schema: v1
models:
  - name: DeepSeek Flash Lite Xploded (MagmaRouter)
    provider: openai
    model: deepseek-v4.1-flash-lite-1017-xploded
    apiBase: https://api.magmarouter.com/v1
    apiKey: ${{ secrets.MAGMAROUTER_API_KEY }}
    defaultCompletionOptions:
      contextLength: 512000
      maxTokens: 32768
    capabilities:
      - tool_use
    roles:
      - chat
      - edit
      - apply
Continue with DeepSeek Flash Lite Xploded Continue only sends tools to models it believes can use them. The capabilities line is that belief; without it agent mode quietly falls back to chat.

DeepSeek Flash Lite Xploded has its own page with every limit and price.

Uncensored DeepSeek V4.1 Flash sets it against the rest of the line, and uncensored DeepSeek goes through the DeepSeek releases one by one.

What a session costs

For a 25-turn agent task (about 750K input tokens re-read across turns, 85% of them cache hits, 25K output), DeepSeek Flash Lite Xploded comes to about $0.096 at today's prices; the cache price does most of the work, since re-read context is billed at $0.030 rather than $0.36 per million.

FAQ

Can DeepSeek Flash Lite Xploded edit files and run commands in Continue?

Yes. Continue acts through tool calls and this model supports tool calling, so it can read, edit and run, not just chat.

What does a session cost?

About $0.096 for a 25-turn agent task (about 750K input tokens re-read across turns, 85% of them cache hits, 25K output), at $0.36 in, $1.44 out and $0.030 per million on cache hits. Put a spend cap on the key so a long run cannot overrun it.

Is it uncensored?

DeepSeek V4.1 Flash 1017 made uncensored with the Heretic method: half a million tokens of context, tool calling, reasoning and vision. The budget DeepSeek for agents.

The other model in Continue