LLM7.io
LLM7.io
Toggle theme

Chat comparison

Anthropic logo

claude-fable-5-2vsclaude-haiku-5-5

Anthropic logo

Compare the things that matter before you build: current price, capabilities, and latest aggregated LLM7 statistics.

Current pricing

See the cost difference clearly

claude-fable-5-2

Checking current pricing…

claude-haiku-5-5

Checking current pricing…

Capability scores

AI-estimated ratings on 30 criteria from head-to-head comparisons against 102 other text models in the LLM7 catalog, judged by TypeSafe Jev from public information (2026-11-19). Each score is the win rate against the field, relative to the strongest model for that criterion (100%). Not a benchmark.

Below 50%50–79%80–94%95%+

claude-fable-5-2 · leads in 27

93%

Strong

claude-haiku-5-5 · leads in 3

68%

Moderate

Knowledge & reasoning

  • General knowledge

    claude-fable-5-2Leads100%
    claude-haiku-5-563%
  • Logical reasoning

    claude-fable-5-2Leads100%
    claude-haiku-5-565%
  • Mathematics

    claude-fable-5-2Leads100%
    claude-haiku-5-560%
  • Scientific and technical depth

    claude-fable-5-2Leads100%
    claude-haiku-5-555%
  • Planning and task decomposition

    claude-fable-5-2Leads100%
    claude-haiku-5-567%

Coding & tools

  • Code generation

    claude-fable-5-2Leads90%
    claude-haiku-5-556%
  • Debugging and code review

    claude-fable-5-2Leads100%
    claude-haiku-5-563%
  • Agentic coding

    claude-fable-5-2Leads100%
    claude-haiku-5-569%
  • Tool and function calling

    claude-fable-5-2Leads100%
    claude-haiku-5-572%
  • SQL and query writing

    claude-fable-5-2Leads100%
    claude-haiku-5-563%

Writing & language

  • Creative writing

    claude-fable-5-2Leads99%
    claude-haiku-5-564%
  • Professional writing

    claude-fable-5-2Leads100%
    claude-haiku-5-572%
  • Conversational quality

    claude-fable-5-2Leads100%
    claude-haiku-5-571%
  • Multilingual ability

    claude-fable-5-2Leads76%
    claude-haiku-5-555%
  • Translation

    claude-fable-5-2Leads100%
    claude-haiku-5-567%
  • Summarization

    claude-fable-5-2Leads100%
    claude-haiku-5-572%

Data & professional work

  • Data analysis

    claude-fable-5-2Leads100%
    claude-haiku-5-567%
  • Professional domains

    claude-fable-5-2Leads100%
    claude-haiku-5-566%
  • Information extraction

    claude-fable-5-2Leads100%
    claude-haiku-5-571%
  • Classification and labeling

    claude-fable-5-2Leads100%
    claude-haiku-5-571%
  • Long-context handling

    claude-fable-5-2Leads100%
    claude-haiku-5-582%

Reliability & control

  • Instruction following

    claude-fable-5-2Leads100%
    claude-haiku-5-572%
  • Structured output

    claude-fable-5-2Leads99%
    claude-haiku-5-580%
  • Factual accuracy

    claude-fable-5-2Leads100%
    claude-haiku-5-577%
  • Safety and alignment

    claude-fable-5-2Leads99%
    claude-haiku-5-577%
  • Steerability

    claude-fable-5-2Leads98%
    claude-haiku-5-576%
  • Output consistency

    claude-fable-5-282%
    claude-haiku-5-5Leads84%

Speed, cost & vision

  • Response speed

    claude-fable-5-215%
    claude-haiku-5-5Leads77%
  • Cost efficiency

    claude-fable-5-219%
    claude-haiku-5-5Leads70%
  • Image understanding

    claude-fable-5-2Leads100%
    claude-haiku-5-542%

Latest LLM7 statistics

Recent aggregated percentages and response-time metrics from LLM7, not a global model benchmark or traffic disclosure.

Recent stability

claude-fable-5-2

98.39%

claude-haiku-5-5

99.24%

Average response time

claude-fable-5-2

6,544.56 ms

claude-haiku-5-5

5,181.92 ms

P95 API latency

claude-fable-5-2

30,000 ms

claude-haiku-5-5

30,000 ms

Side-by-side trends

Shared charts make the observed differences easier to read. Metrics without enough history stay hidden.

Request success rate

Both models on the same scale

claude-fable-5-2: 100%claude-haiku-5-5: 100%
20 Oct19 Nov

Average API latency

Both models on the same scale

claude-fable-5-2: 3,713 msclaude-haiku-5-5: 5,550 ms
20 Oct19 Nov

P95 API latency

Both models on the same scale

claude-fable-5-2: 10,000 msclaude-haiku-5-5: 30,000 ms
20 Oct19 Nov

Estimate your workload

Published rates are loaded from the pricing API; per-request quotes are billed from actual provider usage.

Checking current pricing…
Checking current pricing…

Quick take

  • The recent stability rates are too close to distinguish meaningfully.

Keep exploring