Claude Haiku 5.5
Claude Haiku 5.5Alternatives?
BetterClaude Opus 5unstated+23%+4,900%$5 / $25
OpenKimi K3unstated-4%+2,900%$3 / $15Sampling parameters are fixed+3
BetterGemini 3.1 Pro Previewunstated+15%+2,150%$2 / $12Cannot use both thinking controlsBetterGemini 4 Argonunstated+14%+1,900%*$2 / $10Synchronous output stops at 65,536+2BetterClaude Fable 5unstated+9%+9,900%$10 / $50Claude Opus 5unstated+23%+4,900%$5 / $25Gemini 3.1 Pro Previewunstated+15%+2,150%$2 / $12Cannot use both thinking controlsGemini 4 Argonunstated+14%+1,900%*$2 / $10Synchronous output stops at 65,536+2Claude Fable 5unstated+9%+9,900%$10 / $50Gemini 3 Flash Previewunstated+8%+463%$0.50 / $3Kimi K3unstatedOpen-4%+2,900%$3 / $15Sampling parameters are fixed+3
MiniMax M3unstatedOpen-10%+162%$0.30 / $1.20Context: 1M → 524K+2
MiMo v2.6 ProunstatedOpen-13%+172%$0.43 / $0.87Pass back reasoning_content+3
Hy4 previewOpen-16%+525%$0.83 / $2.50No image input
GLM 5.3unstatedOpen-17%+975%$1.40 / $4.40No image input+4Claude Opus 5unstated+23%—$5 / $25Claude Fable 5unstated+9%—$10 / $50Claude Opus 4.7unstated+6%—$5 / $25Claude Fable 5.1unstated+4%—$10 / $50Replace forced tool_choice+5Claude Opus 4.8unstated+3%—$5 / $25Claude Sonnet 5.5unstated+3%—$2 / $10No forced tool use+5Claude Opus 5.5unstated-3%—$4 / $20No forced tool use+5Claude Opus 4.5thinking-5%—$5 / $25Context: 1M → 200KAll Anthropic models →History
3 versionsBenchmarks
in Chat & KnowledgeWatch out 6
Remove sampling parameters›
Unsupported sampling values, any top_k or combined temperature and top_p return HTTP 400; omit all three and use prompting. Source ↗
Replace manual thinking budgets›
Manual thinking with budget_tokens returns HTTP 400; replace the old thinking configuration. Source ↗
No assistant prefill›
A final assistant prefill returns HTTP 400 even with thinking off; end messages with a user turn. Source ↗
Replace old computer tool · Claude API / Google Cloud›
computer_20250124 returns HTTP 400 on the Claude API and Google Cloud; use computer_toolset_20260801 there. Source ↗
Keep conversations append-only›
Replaying thinking after earlier history changes returns HTTP 400, with older accounts checked only when the binding field is set; keep history append-only. Source ↗
Bedrock needs another computer version · Bedrock›
Bedrock does not accept computer_20250124; use computer_20251124 with the computer-use-2025-11-24 beta header. Source ↗
FAQ
How much does Claude Haiku 5.5 cost?›
$0.10 per 1M input tokens and $0.50 per 1M output tokens; cached input $0.01 per 1M. A typical benchmark task costs about $0.047 at medium effort.
Is Claude Haiku 5.5 free to use?›
There is no free API; the cheapest paid price is $0.10 per 1M input tokens.
How good is Claude Haiku 5.5, and what is it best at?›
#66 of 334 in Chat & Knowledge, #30 of 132 in Coding, #48 of 133 in Agents & Tool Use, #16 of 23 in Writing & Language, #39 of 133 in Vision & Documents. Strongest at: Coding, Reasoning, Knowledge.
Is Claude Haiku 5.5 better than Claude Haiku 4.5?›
On Vals MedCode, Claude Haiku 5.5 scores 51.5 vs 32.7 — higher. It is 83% cheaper per task.
How can I use Claude Haiku 5.5?›
Through the API of Amazon Bedrock, Anthropic, Azure, Claude Platform on AWS, Google.
What is the context window of Claude Haiku 5.5?›
1M tokens, with up to 128K output tokens.
When was Claude Haiku 5.5 released? Is there a newer version?›
Released 2026-10-07. It is the current version in its line.
Is Claude Haiku 5.5 open source?›
No, Claude Haiku 5.5 is a closed model available through an API. The closest open-weights option here is Kimi K3.
Can Claude Haiku 5.5 read images?›
It accepts text and images, PDF files as input.
What is Claude Haiku 5.5?›
Claude Haiku 5.5 is a closed language model from Anthropic, released Oct 2026, the successor to Claude Haiku 4.5.
What are the best alternatives to Claude Haiku 5.5?›
Claude Opus 5 (better), Kimi K3 (open weights).