None
None
Long Story Video Skill
Turn your idea into 10s–10min video—AI writes scripts, prompts, and generates footage automatically.
Ads Video Skill
Generate professional ads and sales videos—AI auto-generates scripts, prompts, and footage.
3D Science Explainer Video Skill
Convert scientific concepts into stunning 3D explain animations
Feedback
freeTrialImage.bannerPity
freeTrialImage.upgradeUnlock
- ✓freeTrialImage.benefitHd
- ✓freeTrialImage.benefitWatermark
- ✓freeTrialImage.benefitUnlimited
Haiku 5.5 Benchmark
Explore the Haiku 5.5 Benchmark: verified launch scores, pricing tiers and context limits, all traced to their sources—no hands-on claims.
All Tools
Discover our comprehensive AI-powered animation toolkit
Lyrics to Song
Turn your lyrics into professional songs with AI
Suno 5.5
Create Professional Music with AI
AI Music Generator
Create music from videos with AI
AI Song Generator
Create songs from videos with AI
Suno AI Music Generator
Create Professional Music with AI
Seedance2.0
The Future of AI Video Is Here.
Happy Horse 1.0
Veo3.1
Create Stunning Videos with Veo3.1
Making Sense of the Haiku 5.5 Benchmark
Launch figures, pricing tiers and effort-based costs, each tied to its configuration and the date it was retrieved.
- How Launch Metrics Differ by ScaleElo, partial-credit, coding and accuracy scores sit on incompatible scales, so averaging them would mislead.
- Cost Efficiency for High-Volume TasksDesigned for extraction, classification and summarization jobs that run thousands of times.
- Two Limits That Reshape the Score TablesLow rates end at 100K tokens; the 72.4% computer-use figure is partial credit, not a strict pass rate.
A Practical Guide to the Haiku 5.5 Benchmark
Three steps to convert published Haiku 5.5 benchmark figures into a workload decision.
Haiku 5.5 Benchmark: What the Data Shows
Specifications, pricing tiers, coding outcomes, computer-use results and third-party index data for Claude Haiku 5.5.
Context Size and Output Limits
Text and image inputs produce text output with adaptive thinking enabled. A 1M-token context window and 128K output cap far exceed Haiku 4.5's 200K and 64K.
Pricing Split by Prompt Length
Above 100K prompt tokens, input, output and caching each move to a tier five times the low-cost rate. OpenAI's base Standard tier extends to 272K input tokens.
Terminal-Bench 4.0 Task Completion
Haiku's 39.2% launch score clearly tops Luna yet remains below published Sol rows. Runs allow an eight-hour agent timeout; 2.1 and 3.0 figures come from an older experiment.
FrontierCode Merge Rate and Effort Spend
Scoring includes correctness, tests, scope, style and repo conventions. Haiku reaches 46.4% at max effort and 41.6% at medium for $0.13 per rollout; moving from xhigh to max doubles spend for half a point.
OSWorld 2.1: Partial Credit vs. Strict Pass
The run uses 82 offline tasks, five attempts each, 1080p screenshots and a 500-step cap. For workflows that must complete, the 37.1% strict pass rate is the honest figure.
Third-Party Index Results and Tool Gains
Artificial Analysis lists Haiku across all five effort levels on Intelligence Index v4.3.2 with rounded task costs. HLE scores 45.9% without tools and 57.4% with them—a lift from the surrounding system.
Haiku 5.5 Benchmark: Common Questions
Answers to frequent questions on Claude Haiku 5.5 benchmark results, pricing tiers and context limits.
Where does Haiku 5.5 stand versus GPT-6 Luna and GPT-6 Sol?
For knowledge work, screenshot tasks and narrow coding, Haiku matches Luna closely. Sol and Sol 6.1 cost more and justify it on coding, while Sonnet 5.5 still leads most rows.
Why doesn't the 72.4% computer-use score equal a success rate?
When every checkpoint must be satisfied, the same run reports just 37.1%. An agent may find and edit the right file, then save it in the wrong format or quit before the final action. Include an end-state check.
At what point does Haiku 5.5 move to the higher price tier?
The lowest rates apply only to prompts up to 100K tokens; beyond that, every category costs five times more. Anthropic's 90%, 50% and ~75% savings figures use different baselines.
Is local deployment possible for Haiku 5.5, and how much GPU memory would it require?
The published specs list no parameter count, active-parameter figure, layer count or training compute. 'Small' describes a product tier, not an architecture, so local GPU estimates lack a basis.
Does higher reasoning effort always lead to better results?
Not always. Sol 6.1 scores higher at medium than at max, and Haiku's 0.2-point lead over Sonnet at max is too small to declare a win. More effort adds tokens, latency and chances to drift off task.
Why do identical models post different numbers across tables?
Sonnet shows 70.6% in Anthropic's launch table and 61.8% on the public leaderboard snapshot, where Haiku had no visible row. Artificial Analysis removed its 'Default Fallback' label, yet Anthropic's migration guide still states no server-side fallback exists.
Put the Haiku 5.5 Benchmark to Work on Your Task
Choose a workload, set the effort level, and measure Haiku 5.5 against your own quality bar.