Anthropic model product
Claude Haiku 4.5
5 is Anthropic's fastest, most cost-efficient model, matching Sonnet 4's performance on coding, computer use, and agent tasks.
Updated Aug 12, 2026. Default version: Claude Haiku 4.5
LLMBoard score45.5Claude Haiku 4.5
Coverage80%12 benchmark families
Context window200KTokens
Official input price$1Anthropic API
Claude Haiku 4.5 Specifications
Technical details for the model's default version.
- Version
- Claude Haiku 4.5
- Released
- Oct 15, 2025
- Knowledge cutoff
- Feb 1, 2025
- Parameters
- N/A
- Context window
- 200K
- Max output
- 64K
- Inputs
- image, pdf, text
- Outputs
- text
- Open weights
- No
- License
- Proprietary
Claude Haiku 4.5 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Claude Haiku 4.5 Benchmark Results
Benchmark scores for Claude Haiku 4.5.
| Cybersecurity CTFs | 46.9% | 02 | 3 | 50.0% | C | |
| MMMU (validation) | 73.2% | 04 | 4 | 0.0% | C | |
| Tau2 Retail | 83.2% | 05 | 26 | 84.0% | C | |
| Terminal-Bench | 41.0% | 07 | 25 | 75.0% | C | |
| OSWorld | 50.7% | 09 | 20 | 57.9% | C | |
| Tau2 Airline | 63.6% | 10 | 23 | 59.1% | C | |
| Tau2 Telecom | 83.0% | 21 | 35 | 41.2% | C | |
| Finance Agent v2 | 31.0% | 23 | 26 | 12.0% | B | |
| MMMLU | 83.0% | 36 | 49 | 27.1% | C | |
| SWE-Bench Verified | 73.3% | 46 | 105 | 56.7% | C | |
| AIME 2025 | 80.7% | 70 | 114 | 38.9% | C | |
| GPQA | 73.0% | 120 | 234 | 48.9% | C |
Claude Haiku 4.5 Arena Results
Preference and agent-evaluation results for the default version.
| document | overall | 30 | 1423.0 | 30,937 | N/A | |
| document style control | overall | 36 | 1422.2 | 30,937 | N/A | |
| webdev | overall | 87 | 1325.7 | 25,722 | N/A | |
| text factuality | overall | 115 | 1416.3 | 111,121 | N/A | |
| text style control | overall | 124 | 1412.2 | 117,833 | N/A | |
| text | overall | 139 | 1393.9 | 117,833 | N/A |
Claude Haiku 4.5 Runtime Performance
Provider-specific output speed and catalog latency for Claude Haiku 4.5. Runtime does not affect the capability score.
| Vertex AI | 100 tok/s | 0.3 s | 200K | 64K | |
| Anthropic | 3.682 tok/s | 0.3 s | 200K | 64K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Claude Haiku 4.5 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $1 input, $5 output per 1M
- Official provider
- Anthropic
- Lowest third-party
- From $0.14 input, $0.71 output per 1M via QiHang
- Tracked offerings
- 45
| QiHang | claude-haiku-4-5-20251001 | global | $0.14 | $0.71 | 200K | |
| Xpersona | claude-haiku-4-5 | global | $0.60 | $3.7 | 200K | |
| Poe | anthropic/claude-haiku-4.5 | global | $0.85 | $4.3 | 192K | |
| Jiekou.AI | claude-haiku-4-5-20251001 | global | $0.90 | $4.5 | 20K | |
| Cortecs | claude-haiku-4-5 | global | $0.996 | $4.98 | 200K | |
| Impossibl | anthropic/claude-haiku-4-5 | global | $1 | $5 | 200K | |
| CrossModel | anthropic/claude-haiku-4-5 | global | $1 | $5 | 200K | |
| NanoGPT | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| NEAR AI Cloud | anthropic/claude-haiku-4-5 | global | $1 | $5 | 200K | |
| DaoXE | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| LLM Gateway | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| LLM Gateway | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| OpenCode Zen | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| GitHub Copilot | claude-haiku-4.5 | global | $1 | $5 | 200K | |
| 302.AI | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| 302.AI | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| Helicone | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| Neon | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| FrogBot | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| Pioneer | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| ZenMux | anthropic/claude-haiku-4.5 | global | $1 | $5 | 200K | |
| Abacus | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| Azure Cognitive Services | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| DigitalOcean | anthropic-claude-4.5-haiku | global | $1 | $5 | 200K | |
| Azure | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| Vertex (Anthropic) | claude-haiku-4-5@20251001 | global | $1 | $5 | 200K | |
| Ofox | anthropic/claude-haiku-4.5 | global | $1 | $5 | 200K | |
| OrcaRouter | anthropic/claude-haiku-4.5 | global | $1 | $5 | 200K | |
| Vertex | claude-haiku-4-5@20251001 | global | $1 | $5 | 200K | |
| FreeModel | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| Amazon Bedrock | anthropic.claude-haiku-4-5-20251001-v1:0 | global | $1 | $5 | 200K | |
| Requesty | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| Kilo Gateway | anthropic/claude-haiku-4.5 | global | $1 | $5 | 200K | |
| Merge Gateway | anthropic/claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| Vercel AI Gateway | anthropic/claude-haiku-4.5 | global | $1 | $5 | 200K | |
| OpenRouter | anthropic/claude-haiku-4.5 | global | $1 | $5 | 200K | |
| Anthropic | claude-haiku-4-5 | global | $1 | $5 | 200K | |
| Anthropic | claude-haiku-4-5-20251001 | global | $1 | $5 | 200K | |
| Cloudflare AI Gateway | anthropic/claude-haiku-4-5 | global | $1 | $5 | 200K | |
| Perplexity Agent | anthropic/claude-haiku-4-5 | global | $1 | $5 | 200K | |
| Databricks | databricks-claude-haiku-4-5 | global | $1 | $5 | 200K | |
| UnoRouter | claude-haiku-4-5-20251001 | global | $1.2 | $6 | 200K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Claude Haiku 4.5 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Claude Haiku 4.5 | 45.5 | N/A | 200K | 64K | No | Proprietary |
What is Claude Haiku 4.5?
Key information about Claude Haiku 4.5 and its available data.
5 is Anthropic's fastest, most cost-efficient model, matching Sonnet 4's performance on coding, computer use, and agent tasks. It offers similar performance to Sonnet 4 at one-third the cost and more than twice the speed, making it ideal for high-volume, latency-sensitive applications and multi-agent orchestration.
Data as of 2026-08-11.
Claude Haiku 4.5 vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Claude Haiku 4.5
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AN
Claude Opus 4
Anthropic
AN
Claude Sonnet 4
Anthropic
AN
Claude Sonnet 3.7
Anthropic
AN
Claude Opus 4.1
Anthropic
AN
Claude Sonnet 4.5
Anthropic
AN
Claude Sonnet 3.5
Anthropic
FAQ
Common questions about Claude Haiku 4.5.
When was Claude Haiku 4.5 released?
Claude Haiku 4.5's default version was released on Oct 15, 2025.
How much does Claude Haiku 4.5 cost?
Claude Haiku 4.5's official API price is $1 per million input tokens and $5 per million output tokens via Anthropic. The lowest tracked third-party offer starts at $0.14 input and $0.71 output via QiHang.
Who created Claude Haiku 4.5?
Claude Haiku 4.5 was created by Anthropic.
What is the context window for Claude Haiku 4.5?
The default version has a 200K token context window.
Is Claude Haiku 4.5 open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer Claude Haiku 4.5?
45 provider offerings are linked to the default version.
What models should I compare Claude Haiku 4.5 with?
Nearby ranked alternatives include GPT-OSS-20B, Claude Opus 4, Qwen3 VL 235B A22B Thinking.