Claude Haiku 4.5 (Fast)
LLM Provider / Model · High-throughput coding sidekick, autocomplete, batch jobs, and low-latency Claude Code use
At a glance
| Input price | $1 |
| Output price | $5 |
| Cache price | $0.10 |
| Price tier | Budget |
| Context window | 200K |
| Max output | 32K |
| Context tier | 128K-500K |
| Speed tier | Fast |
| Latency | low |
| Knowledge cutoff | Oct 2025 |
| Modality | Multimodal (vision) |
| Model ID | claude-haiku-4-5-fast |
| Provider | Anthropic |
| HumanEval | 88% |
| MMLU | 82% |
| SWE-Bench | 55% |
| Benchmark | Best-in-class latency at Haiku-tier intelligence |
| Released | Oct 2025 |
| Hosting | Closed/API |
| Capabilities | Vision, Tool use, Streaming, Structured output, Prompt caching, Low-latency mode |
What Claude Haiku 4.5 (Fast) does
Vision, Tool use, Streaming, Structured output, Prompt caching, Low-latency mode
Best for
High-throughput coding sidekick, autocomplete, batch jobs, and low-latency Claude Code use
Works well with
Integration
Agent / Orchestration
Conflicts & caveats
- OpenAI Codex (VS Code) is tied to OpenAI's GPT-5.1 Codex family — Claude Haiku 4.5 (Fast) is not selectable inside this integration.
- Gemini CLI is locked to Google Gemini models — your LLM is Claude Haiku 4.5 (Fast).
- Devin uses Cognition's proprietary model internally — your selected Claude Haiku 4.5 (Fast) will be ignored.
- OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected Claude Haiku 4.5 (Fast) will be ignored.
- Jules runs Gemini models internally — your selected Claude Haiku 4.5 (Fast) will be ignored.
- Lovable uses a fixed internal model — your selected Claude Haiku 4.5 (Fast) will be ignored.
- Replit Agent 3 uses its own model selection internally — your selected Claude Haiku 4.5 (Fast) will be ignored.
Build a full stack around Claude Haiku 4.5 (Fast) — Flowpicker shows compatibility warnings before you commit.
Open the stack planner →