Results / stream.utf8-split
Stream chunks split inside UTF-8 characters
The same bytes as stream.utf8 are split into chunks whose boundaries fall inside multi-byte characters.
Bug class: Streamed content or completion state corrupted
The library corrupts streamed text, for example at multi-byte character boundaries, or reports a stream that ended early as complete.
Corrupted or unfinished text is treated as the final answer.
Requirements
| Level | Id | Check | Requirement | Providers |
|---|---|---|---|---|
| MUST | text-reported | text_matches_truth | The streamed text is reported exactly as the provider sent it. | All |
| MUST | no-replacement-chars | no_replacement_chars | The reported text contains no replacement characters. | All |
Fixtures
| Provider | Tier | any-llm | litellm | pydantic-ai | tanstack-ai | vercel-ai |
|---|---|---|---|---|---|---|
| anthropic | derived | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS |
| gemini | derived | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS |
| ollama | derived | ✓ PASS | ✓ PASS | – UNSUPPORTED | ✓ PASS | – UNSUPPORTED |
| ollama-openai | derived | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS |
| openai-chat | derived | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS |
| openai-responses | derived | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS | ✓ PASS |