Skip to main content
Van Data Team

text model comparison

Claude Fable 5.2 vs GPT-6 Astra

Claude Fable 5.2 and GPT-6 Astra are both current text models, from Anthropic and OpenAI. Claude Fable 5.2 is 20% cheaper per 1M input tokens than GPT-6 Astra, $8 vs $10. On GPQA Diamond, Claude Fable 5.2 scores 93.9% and GPT-6 Astra 95.8% (Epoch AI, measured November 10, 2026). GPT-6 Astra has the larger context window, 1,050,000 vs 1,000,000 tokens.

Higher GPQA Diamond
GPT-6 Astra95.8% against 93.9%
Lower price per 1M input tokens
Claude Fable 5.2$8 against $10
Longer context window
GPT-6 Astra1.05M tokens against 1M tokens
Newer release
Claude Fable 5.2Nov 2026 against Sep 2026

Data verified November 10, 2026 · Dataset 2026-11-10

Share this

Side by side

Claude Fable 5.2 and GPT-6 Astra, line by line

Prices in effect today, compared at the same tier wherever both list one. The better value in each row is highlighted.

FactClaude Fable 5.2GPT-6 Astra
Price
Price per 1M input tokens$8 (better)$10
Price per 1M output tokens$40 (better)$50
Independent benchmarks
GPQA Diamond93.9% (#12 of 43)95.8% (#2 of 43) (better)
FrontierMath Tiers 1-3 (v2)95.1% (#1 of 43) (better)93.7% (#2 of 43)
SimpleQA Verified94% (#1 of 40) (better)75.6% (#3 of 40)
Arena TextNot measured1474 Elo (#22 of 53)
Arena WebDevNot measured1767 Elo (#4 of 48)
Specs
ProviderAnthropicOpenAI
Release dateNovember 9, 2026September 3, 2026
AccessAPI onlyAPI only
Context window1,000,000 tokens1,050,000 tokens
Max output128,000 tokens128,000 tokens
Inputtext, imagetext, image
Outputtexttext

Scores:Epoch AI · 2026-11-10Arena · 2026-11-10

Open both in the interactive comparison

Provider results

What the labs report

Results tables a lab published with one of its models, where it lists both Claude Fable 5.2 and GPT-6 Astra. These are the lab's own figures, not independent measurements, and each table is read on its own: labs run benchmarks with different harnesses and budgets.

Anthropic, Claude Fable 5.2 launch table, November 9, 2026Anthropic · 2026-11-10

BenchmarkClaude Fable 5.2GPT-6 Astra
FrontierCode 2.0best effort level per model49.1% (better)42.8%
FrontierSWE 269.9% (better)65.5%
AutomationBenchZapier51.4% (better)41.4%
Terminal-Bench-Science 0.174.6% (better)68.1%
OSWorld 2.1offline set, partial credit89.3% (better)79.5%
Remote Labor Index41.3% (better)20.8%
Humanity's Last Examwith tools74.2% (better)57.2%
Chartographyno tools73.4% (better)71.0%

Anthropic's table from the Claude Fable 5.2 and Mythos 5.2 announcement, each model at its highest available effort, with Anthropic's safeguards on; competitor figures come from their developers' system cards or the benchmarks' leaderboards. On FrontierCode 2.0 Fable 5.2 and 5.1 ran at xhigh effort and Gemini 4 Carbon at medium. Anthropic gives a Mythos 5.2 score for Terminal-Bench-Science only. Left out here: Anthropic's own evaluations (CVE Rediscovery, de novo binder design, RoboDojo Sim), Irregular's CyScenarioBench, where each lab's score covers different tasks, and Artificial Analysis's AA-Briefcase.

Charts

Against every current text model

Where Claude Fable 5.2 and GPT-6 Astra land among the current models of their task, on every independent benchmark, on price and on context.

Where each stands

On GPQA Diamond, Claude Fable 5.2 ranks #12 and GPT-6 Astra #2 of 43 current models. Each bar is the share of current models it beats: on score, price and context.

Claude Fable 5.2 and GPT-6 Astra among the current text models, one row per benchmark, price and context window.

Independent benchmarks

Arena Text

Claude Fable 5.2: Not scored

GPT-6 Astra:#22of 53

1,474 Elo · ahead of 58%

Arena WebDev

Claude Fable 5.2: Not scored

GPT-6 Astra:#4of 48

1,767 Elo · ahead of 94%

GPQA Diamond

Claude Fable 5.2:#12of 43

93.9% · ahead of 71%

GPT-6 Astra:#2of 43

95.8% · ahead of 98%

FrontierMath Tiers 1-3 (v2)

Claude Fable 5.2:#1of 43

95.1% · ahead of 100%

GPT-6 Astra:#2of 43

93.7% · ahead of 95%

SimpleQA Verified

Claude Fable 5.2:#1of 40

94% · ahead of 100%

GPT-6 Astra:#3of 40

75.6% · ahead of 95%

Price and context

Price per 1M input tokens

Claude Fable 5.2:#53of 61

$8 · cheaper than 12%

GPT-6 Astra:#55of 61

$10 · cheaper than 5%

Price per 1M output tokens

Claude Fable 5.2:#53of 61

$40 · cheaper than 12%

GPT-6 Astra:#55of 61

$50 · cheaper than 5%

Context window

Claude Fable 5.2:#28of 70

1M tokens · longer than 23%

GPT-6 Astra:#1of 70

1.05M tokens · longer than 86%

Scores fromArena · 2026-11-10Epoch AI · 2026-11-10

Quality against price

Arena Text · measured November 9, 2026

GPT-6 Astra scores 1,474 at $20 (blended price per 1M tokens). The line joins the current models no other beats on both price and Arena Text score; hover any point to name it.

Quality against priceArena Text score against price for 45 text models, Claude Fable 5.2 and GPT-6 Astra marked.1,2501,3001,3501,4001,4501,5001,550$0.30$1$3$10MOST ATTRACTIVE QUADRANTMedian price $3Median 1,473Blended price per 1M tokens (3 input : 1 output), log scaleArena Text (Elo)GPT-6 Astra
Quality against price: the data
Arena Text score against price for 45 text models, Claude Fable 5.2 and GPT-6 Astra marked.
ModelProviderBlended price per 1M tokensArena Text (Elo)
Gemini 4 CarbonGoogle$41,526
Gemini 4 ArgonGoogle$41,522
Muse WonderMeta$81,511
Claude Opus 5.5Anthropic$81,506
Claude Fable 5Anthropic$201,504
Claude Fable 5.1Anthropic$201,503
Gemini 3.8 FlashGoogle$1.501,497
Muse Spark 1.3Meta$21,496
Claude Opus 5Anthropic$101,490
Kimi K3Moonshot Kimi$61,487
GPT-5.6 SolOpenAI$81,485
Qwen3.8-MaxAlibaba Qwen$31,484
Gemini 3.6 FlashGoogle$31,483
GLM-5.3Z.ai$2.151,483
Claude Opus 4.8Anthropic$101,482
GPT-6.1 SolOpenAI$41,482
GPT-5.5OpenAI$111,481
MiMo-V2.6-ProXiaomi MiMo$0.541,480
GPT-5.4OpenAI$5.631,475
DeepSeek-V4.1-FlashDeepSeek$0.521,475
GPT-6 AstraOpenAI$201,474
GLM-5.3-FlashZ.ai$0.241,474
Gemini 3 Flash PreviewGoogle$1.131,473
Grok 4.20xAI$1.561,472
Claude Sonnet 5.5Anthropic$41,472
GPT-5.6 TerraOpenAI$4.501,468
Grok 4.5xAI$31,466
DeepSeek-V4-ProDeepSeek$1.981,464
Claude Sonnet 5Anthropic$41,461
GPT-6 SolOpenAI$41,457
Grok 4.6xAI$31,456
Gemini 3.5 Flash-LiteGoogle$0.851,455
GPT-5.6 LunaOpenAI$0.451,453
MiMo-V2.6-FlashXiaomi MiMo$0.181,448
Step 5 PreviewStepFun$1.431,448
Grok 4.3xAI$1.561,442
Grok 4.7xAI$31,442
GPT-6 LunaOpenAI$0.201,441
Mistral Large 4Mistral AI$1.051,423
Claude Haiku 5.5Anthropic$0.201,422
GPT-4.1OpenAI$3.501,415
Inkling 2 SmallThinking Machines Lab$0.471,412
Solar Pro 4Upstage$0.521,387
GPT-4.1 miniOpenAI$0.701,383
GPT-4.1 nanoOpenAI$0.181,322

Cost

What each costs

Example workloads at each provider's listed rates, then every other rate either one lists. The cheaper figure in each row is highlighted.

The same work, priced

For 10M input and 2M output tokens, GPT-6 Astra costs 1.3 times as much as Claude Fable 5.2, at each provider's listed rates.

What Claude Fable 5.2 and GPT-6 Astra cost for the same work, at the rates each provider lists.

  • 10M input and 2M output tokens

    GPT-6 Astra costs 1.3× as much

    Claude Fable 5.2
    $160
    GPT-6 Astra
    $200

  • The same, with 9M of the input tokens cached

    GPT-6 Astra costs 1.3× as much

    Claude Fable 5.2
    $89.80
    GPT-6 Astra
    $119
Workload or rateClaude Fable 5.2GPT-6 Astra
Example workloads
10M input and 2M output tokens$160 (better)$200
The same, with 9M of the input tokens cached$89.80 (better)$119
The same, through the batch API$80Not listed
Other listed rates
1M input tokens, cached$0.20 (better)$1
1M input tokens, 5-minute cache write$10Not listed
1M input tokens, 1-hour cache write$16Not listed
1M input tokens, batch$4Not listed
1M output tokens, batch$20Not listed

Your own bill will differ. These figures use list prices and example workloads. The AI Model Cost Audit reviews your real usage, shortlists models that score as well for each task for less, and estimates the savings.

Launch price $100, regular price $150

Book the cost audit

Verdict

When to choose each

Claude Fable 5.2

Where it leads

  • Lower input and output prices, $8 and $40 vs $10 and $50 per 1M tokens
  • Higher FrontierMath Tiers 1-3 (v2) score, 95.1% vs 93.7%
  • Higher SimpleQA Verified score, 94% vs 75.6%
  • Cheaper cached input, $0.20 vs $1 per 1M tokens
  • Batch pricing, $4 input and $20 output per 1M tokens

Best for

  • The hardest open-ended coding, research and agent runs, where Claude Opus 5.5 at high effort falls short
  • Advising or orchestrating cheaper Claude models in a multi-agent setup

Watch out for

  • Gemini 4 Carbon scores higher on AutomationBench in Anthropic's own table (53.9% against 51.4%)
  • Safety classifiers block most penetration testing, exploit development and research biology requests and answer them with an older model (Claude Opus 4.8 for cybersecurity, Claude Opus 5 for biology); vetted organisations use Claude Mythos 5.2 for that work
  • Anthropic recommends starting most coding and knowledge work on Claude Opus 5.5, at half the price, and moving up to Fable only when it falls short; Arena has not rated Fable 5.2 yet
Full Claude Fable 5.2 profile

GPT-6 Astra

Where it leads

  • Higher GPQA Diamond score, 95.8% vs 93.9%
  • Larger context window 1,050,000 vs 1,000,000 tokens

Best for

  • Frontier reasoning and agentic evaluations where deep-context retrieval accuracy is the binding constraint
  • Workloads that can be kept under the long-context threshold, where the effective rate stays at $10 / $50

Watch out for

  • Long-context billing is a cliff, not a taper: a request over the long-context threshold is rebilled in full at $20 per million input and $75 per million output, so a prompt that crosses it costs roughly double
  • GPT-6 Astra is the first publicly released model OpenAI classifies as crossing the Critical cybersecurity threshold under its own Preparedness Framework, which carries additional deployment safeguards
  • Vendor-reported benchmark figures come from OpenAI’s own launch material and its own harness; they are not independently reproduced here
Full GPT-6 Astra profile

From benchmark to production

Building on Claude Fable 5.2 or GPT-6 Astra?

Public scores rarely hold on your own data. We test both models on your real tasks, then build the agent or pipeline around the one that wins.

AI Model Cost AuditLaunch price $100, regular price $150What the audit includes

Book the cost auditAI agent development

FAQ

Claude Fable 5.2 vs GPT-6 Astra FAQ

Which is cheaper, Claude Fable 5.2 or GPT-6 Astra?

Claude Fable 5.2 is 20% cheaper per 1M input tokens than GPT-6 Astra, $8 vs $10. For 10M input and 2M output tokens, Claude Fable 5.2 costs $160 and GPT-6 Astra $200 at standard rates.

Which is cheaper with prompt caching, Claude Fable 5.2 or GPT-6 Astra?

Claude Fable 5.2 charges $0.20 per 1M cached input tokens, 98% below its standard input rate, and GPT-6 Astra $1, 90% below. With 9M of 10M input tokens cached and 2M output tokens, Claude Fable 5.2 costs $89.80 and GPT-6 Astra $119.

Which scores higher on independent benchmarks, Claude Fable 5.2 or GPT-6 Astra?

GPT-6 Astra leads on GPQA Diamond, 95.8% to 93.9%; Claude Fable 5.2 leads on FrontierMath Tiers 1-3 (v2), 95.1% to 93.7%; Claude Fable 5.2 leads on SimpleQA Verified, 94% to 75.6%. Scores are independent evaluations, not either vendor's launch figures.

How do Claude Fable 5.2 and GPT-6 Astra compare in Anthropic's own results?

In the Claude Fable 5.2 launch table (November 9, 2026), Claude Fable 5.2 leads on FrontierCode 2.0 (best effort level per model), 49.1% to 42.8%; Claude Fable 5.2 leads on FrontierSWE 2, 69.9% to 65.5%; Claude Fable 5.2 leads on AutomationBench (Zapier), 51.4% to 41.4%. These are Anthropic's figures, not independent measurements.

Which has the larger context window, Claude Fable 5.2 or GPT-6 Astra?

GPT-6 Astra has the larger context window, 1,050,000 vs 1,000,000 tokens. Both return up to 128,000 tokens per response.

Further reading

More on Claude Fable 5.2 and GPT-6 Astra

Articles on Claude Fable 5.2 and GPT-6 Astra

Related comparisons

Production evaluation

Choosing between Claude Fable 5.2 and GPT-6 Astra?

Public benchmarks narrow the shortlist. We run both on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.

  • A head-to-head evaluation on your real workload
  • Routing between models by cost and difficulty
Compare all modelsBook a consultation