|
1 | 1 | --- |
2 | | -title: "How We Test TanStack AI Across 7 Providers on Every PR" |
| 2 | +title: 'How We Test TanStack AI Across 7 Providers on Every PR' |
3 | 3 | published: 2026-04-13 |
4 | 4 | draft: true |
5 | 5 | excerpt: "TanStack AI runs 147 deterministic E2E tests across 7 LLM providers in under 2 minutes. Here's the testing infrastructure that makes it possible." |
@@ -55,16 +55,16 @@ Every test gets a unique `X-Test-Id` header. This is what makes parallel executi |
55 | 55 |
|
56 | 56 | 147 tests cover 17 features across 7 providers. Here's the matrix: |
57 | 57 |
|
58 | | -| Category | Tests | Features | |
59 | | -|---|---|---| |
60 | | -| Chat and text | 28 | chat, one-shot-text, multi-turn, structured-output | |
61 | | -| Tool calling | 38 | single, parallel, approval, text-tool-text, agentic-structured | |
62 | | -| Multimodal | 10 | image input, structured multimodal | |
63 | | -| Generation | 20 | summarize, summarize-stream, image-gen, TTS, transcription | |
64 | | -| Reasoning | 3 | OpenAI, Anthropic, Gemini thinking blocks | |
65 | | -| Tools-test page | 29 | client tools, approvals, race conditions, server-client sequences | |
66 | | -| Advanced | 6 | abort, lazy tools, custom events, middleware, error handling | |
67 | | -| Middleware | 3 | onChunk transform, onBeforeToolCall skip, passthrough | |
| 58 | +| Category | Tests | Features | |
| 59 | +| --------------- | ----- | ----------------------------------------------------------------- | |
| 60 | +| Chat and text | 28 | chat, one-shot-text, multi-turn, structured-output | |
| 61 | +| Tool calling | 38 | single, parallel, approval, text-tool-text, agentic-structured | |
| 62 | +| Multimodal | 10 | image input, structured multimodal | |
| 63 | +| Generation | 20 | summarize, summarize-stream, image-gen, TTS, transcription | |
| 64 | +| Reasoning | 3 | OpenAI, Anthropic, Gemini thinking blocks | |
| 65 | +| Tools-test page | 29 | client tools, approvals, race conditions, server-client sequences | |
| 66 | +| Advanced | 6 | abort, lazy tools, custom events, middleware, error handling | |
| 67 | +| Middleware | 3 | onChunk transform, onBeforeToolCall skip, passthrough | |
68 | 68 |
|
69 | 69 | **Providers:** OpenAI, Anthropic, Gemini, Ollama, Groq, Grok, OpenRouter. |
70 | 70 |
|
|
0 commit comments