fix(aiService): improve deepseek-reasoner compatibility

[AmintaCCCP]

Closes

Fixes #70

Overview

This PR fixes compatibility issues when using DeepSeek's official deepseek-reasoner model through the OpenAI-compatible chat completions path.

Changes

• detect deepseek-reasoner specifically in the OpenAI-compatible branch
• omit unsupported temperature for that model
• omit OpenAI-specific reasoning payload for that model
• keep all existing behavior unchanged for other model channels

Why this is safe

This change is scoped only to apiType === "openai" with model name exactly deepseek-reasoner. It does not affect:

• other OpenAI-compatible models
• openai-responses
• Claude
• Gemini

Validation

• local frontend build passed with npm run build

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR fixes a compatibility issue with DeepSeek's deepseek-reasoner model by adding detection logic and conditionally omitting request fields (temperature and reasoning) that are incompatible with this specific model variant.

Changes

Cohort / File(s)	Summary
DeepSeek Reasoner Model Compatibility src/services/aiService.ts	Added private helper method isDeepSeekReasonerModel() to detect when the model is deepseek-reasoner. Modified OpenAI chat-completions request construction to conditionally omit temperature and reasoning fields when this model is detected, addressing API compatibility constraints.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~5 minutes

Possibly related PRs

• feat(ai): 新增 OpenAI Responses API 支持（保持现有兼容） #44: Both PRs modify OpenAI request construction in src/services/aiService.ts; the related PR handles "openai-responses" API type branching while this PR adds DeepSeek Reasoner model detection and conditional field handling.

Poem

🐰 A reasoner model, deep and wise,
Needed special care to recognize,
We hopped in fast, conditions tight,
Trimmed fields away—now working right! ✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly describes the main change: improving deepseek-reasoner compatibility in the aiService, which matches the core objective of fixing compatibility issues with DeepSeek's deepseek-reasoner model.
Linked Issues check	✅ Passed	The code changes directly address issue #70 by detecting the deepseek-reasoner model and omitting unsupported parameters (temperature and reasoning), ensuring proper compatibility with DeepSeek's API requirements.
Out of Scope Changes check	✅ Passed	All changes are strictly scoped to deepseek-reasoner compatibility via the OpenAI-compatible path, with no modifications to other model channels, maintaining existing behavior for Claude, Gemini, and other OpenAI-compatible models.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix-deepseek-reasoner-compat

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

src/services/aiService.ts (1)

72-88: Add regression tests for request-body shape in this branch.

Please add a focused test asserting that temperature and reasoning are omitted for deepseek-reasoner, and still present for non-DeepSeek models. This will protect this compatibility fix from future regressions.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/services/aiService.ts` around lines 72 - 88, Add focused unit tests for
the request body shape produced by the code that builds requestBody in
aiService.ts: call the method (or class) that constructs the request body
(exercise isDeepSeekReasonerModel() via setting this.config.model or mocking it)
for both apiType branches ('openai-responses' and others) and assert that when
model is a deepseek-reasoner variant temperature and reasoning are omitted from
the resulting request object, and that for a non-DeepSeek model both temperature
and reasoning are present with the expected values; target the functions
isDeepSeekReasonerModel() and the requestBody construction logic so the tests
fail on regressions and mock options.temperature/options.maxTokens/reasoning as
needed.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@src/services/aiService.ts`:
- Around line 72-88: Add focused unit tests for the request body shape produced
by the code that builds requestBody in aiService.ts: call the method (or class)
that constructs the request body (exercise isDeepSeekReasonerModel() via setting
this.config.model or mocking it) for both apiType branches ('openai-responses'
and others) and assert that when model is a deepseek-reasoner variant
temperature and reasoning are omitted from the resulting request object, and
that for a non-DeepSeek model both temperature and reasoning are present with
the expected values; target the functions isDeepSeekReasonerModel() and the
requestBody construction logic so the tests fail on regressions and mock
options.temperature/options.maxTokens/reasoning as needed.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 45d88cd0-9304-446b-b2cc-bed477c050ea

📥 Commits

Reviewing files that changed from the base of the PR and between ee1b9c6 and 6e7aed5.

📒 Files selected for processing (1)

• src/services/aiService.ts

[AmintaCCCP]

This fix has been included in v0.3.0 release.

The DeepSeek Reasoner compatibility issue has been resolved by removing unsupported temperature and reasoning parameters from the request when using deepseek-reasoner model.

Thank you for the contribution!

🦞