Skip to main content

text model comparison

Claude Fable 5.2 vs GPT-6.2 Astra

Claude Fable 5.2 and GPT-6.2 Astra are both current text models, from Anthropic and OpenAI. Both cost $8 per 1M input tokens. In OpenAI's GPT-6.2 Astra and Sol launch charts, Claude Fable 5.2 scores 72.8% on Agents' Last Exam v1.1 and GPT-6.2 Astra 77.6%. GPT-6.2 Astra has the larger context window, 1,050,000 vs 1,000,000 tokens.

Longer context window
GPT-6.2 Astra1.05M tokens against 1M tokens
Newer release
GPT-6.2 AstraNov 17, 2026 against Nov 9, 2026

Data verified November 18, 2026 · Dataset 2026-11-18

Share this

Side by side

Claude Fable 5.2 and GPT-6.2 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-6.2 Astra
Price
Price per 1M input tokens$8$8
Price per 1M output tokens$40$40
Independent benchmarks
Arena WebDev1826 Elo (#1 of 46)Range 1807 to 1845Not measured
Epoch Capabilities Index173.4 points (#1 of 43)Range 169.9 to 177.5Not measured
GPQA Diamond93.9% (#13 of 40)Not measured
FrontierMath Tiers 1-3 (v2)95.1% (#1 of 39)Not measured
SimpleQA Verified94% (#1 of 36)Not measured
Specs
ProviderAnthropicOpenAI
Release dateNovember 9, 2026November 17, 2026
AccessAPI onlyAPI only
Context window1,000,000 tokens1,050,000 tokens
Max output128,000 tokens128,000 tokens
Inputtext, imagetext, image
Outputtexttext

Ranges under the scores are the publishers' own: published rating ranges from Arena and 90% credible intervals from Epoch AI. Where two ranges overlap, the gap is inside the measurement's own uncertainty, so the row says too close to call and neither score is highlighted or counted as a lead.

No independent benchmark has scored both models yet. Each rank counts the current models scored on that benchmark.

Scores:Arena · 2026-11-18Epoch AI · 2026-11-16

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-6.2 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.

OpenAI, GPT-6.2 Astra and Sol launch charts, November 17, 2026OpenAI · 2026-11-18

BenchmarkClaude Fable 5.2GPT-6.2 Astra
Agents' Last Exam v1.172.8%77.6% (better)
FrontierCode 2.045.6% (better)44.2%
BenchCAD 2.1 Core83.5% (better)82.5%
AutomationBench 1.0.651.4%52.4% (better)

OpenAI's charts from the GPT-6.2 announcement, each model at its highest effort in them: max for the OpenAI and Anthropic models, high for Gemini 4 Carbon. At xhigh, GPT-6.2 Astra scores 78.3% on Agents' Last Exam and Claude Fable 5.2 49.1% on FrontierCode 2.0. OpenAI notes that the Agents' Last Exam leaderboard result for Fable 5.2 leaves out nine tasks that used a fallback model; its Terminal-Bench-Science figure for Fable 5.2 comes from Artificial Analysis, so it is left blank here (Anthropic reports 74.6%). Left out here: the Artificial Analysis Intelligence Index and Harvey-LAB-AA, which Artificial Analysis runs, the Vals Science Index, GDPval v2, SecureSWE and ExploitBench, which OpenAI runs internally, LifeVal and PatchBench, which we could not confirm are public, DeepBinary, which charts running totals of API cost and bug-bounty value across twelve files, and the alignment charts, where lower is better.

Charts

Against every current text model

Where Claude Fable 5.2 and GPT-6.2 Astra land among the current models of their task, on every independent benchmark, on price and on context.

Where each stands

Only Claude Fable 5.2 has a current Arena WebDev score: #1 of 46. Each bar is the share of current models it beats.

Claude Fable 5.2 and GPT-6.2 Astra among the current text models, one row per benchmark, price and context window.

Independent benchmarks

Arena WebDev

Claude Fable 5.2:#1of 46

1,826 Elo · ahead of 100%

Published rating range 1,807 to 1,845

GPT-6.2 Astra: Not scored

Epoch Capabilities Index

Claude Fable 5.2:#1of 43

173.4 · ahead of 100%

90% credible interval 169.9 to 177.5

GPT-6.2 Astra: Not scored

GPQA Diamond

Claude Fable 5.2:#13of 40

93.9% · ahead of 69%

GPT-6.2 Astra: Not scored

FrontierMath Tiers 1-3 (v2)

Claude Fable 5.2:#1of 39

95.1% · ahead of 100%

GPT-6.2 Astra: Not scored

SimpleQA Verified

Claude Fable 5.2:#1of 36

94% · ahead of 100%

GPT-6.2 Astra: Not scored

Price and context

Price per 1M input tokens

Claude Fable 5.2:#57of 64

$8 · cheaper than 8%

GPT-6.2 Astra:#57of 64

$8 · cheaper than 8%

Price per 1M output tokens

Claude Fable 5.2:#57of 64

$40 · cheaper than 8%

GPT-6.2 Astra:#57of 64

$40 · cheaper than 8%

Context window

Claude Fable 5.2:#37of 75

1M tokens · longer than 20%

GPT-6.2 Astra:#1of 75

1.05M tokens · longer than 81%

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

At each provider's listed rates.

What Claude Fable 5.2 and GPT-6.2 Astra cost for the same work, at the rates each provider lists.
  • 10M input and 2M output tokens

    The same price

    Claude Fable 5.2
    $160
    GPT-6.2 Astra
    $160
  • The same, with 9M of the input tokens cached

    GPT-6.2 Astra costs 1.0× as much

    Claude Fable 5.2
    $89.80
    GPT-6.2 Astra
    $91.60
  • The same, through the batch API

    The same price

    Claude Fable 5.2
    $80
    GPT-6.2 Astra
    $80
Workload or rateClaude Fable 5.2GPT-6.2 Astra
Example workloads
10M input and 2M output tokens$160$160
The same, with 9M of the input tokens cached$89.80 (better)$91.60
The same, through the batch API$80$80
Other listed rates
1M input tokens, cached$0.20 (better)$0.40
1M input tokens, 5-minute cache write$10Not listed
1M input tokens, 1-hour cache write$16Not listed
1M input tokens, batch$4$4
1M output tokens, batch$20$20
1M input tokens, prompts over 272K tokensNot listed$16
1M output tokens, prompts over 272K tokensNot listed$60

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

See the cost audit

Verdict

When to choose each

Claude Fable 5.2

Where it leads

  • Cheaper cached input, $0.20 vs $0.40 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 Text has not rated Fable 5.2 yet, and its WebDev rating rests on 1,445 votes so far

GPT-6.2 Astra

Where it leads

  • Larger context window 1,050,000 vs 1,000,000 tokens

Best for

  • Long agentic coding, terminal and science work that ran on GPT-6 Astra, at a lower cost per task
  • Slides, spreadsheets and documents that must follow a template or a profession's conventions

Watch out for

  • A request over 272,000 input tokens is billed in full at $16 input and $60 output per 1M tokens, and a 30-minute cache write costs $10 per 1M tokens, more than plain input
  • OpenAI rates it Critical for cybersecurity, High for biological and chemical capabilities and, the first of its released models to reach that level, High for AI self-improvement under its Preparedness Framework, so the strictest safeguards apply
  • Claude Fable 5.2 and Claude Opus 5.5 still score higher on FrontierCode 2.0 in OpenAI's own chart (45.6% and 45.5% against 44.2% at max), and Arena and Epoch AI have not rated it yet

From benchmark to production

Building on Claude Fable 5.2 or GPT-6.2 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 $150

FAQ

Claude Fable 5.2 vs GPT-6.2 Astra FAQ

Which is cheaper, Claude Fable 5.2 or GPT-6.2 Astra?

Both cost $8 per 1M input tokens. For 10M input and 2M output tokens, Claude Fable 5.2 costs $160 and GPT-6.2 Astra $160 at standard rates.

Which is cheaper with prompt caching, Claude Fable 5.2 or GPT-6.2 Astra?

Claude Fable 5.2 charges $0.20 per 1M cached input tokens, 98% below its standard input rate, and GPT-6.2 Astra $0.40, 95% below. With 9M of 10M input tokens cached and 2M output tokens, Claude Fable 5.2 costs $89.80 and GPT-6.2 Astra $91.60.

How do Claude Fable 5.2 and GPT-6.2 Astra compare in OpenAI's own results?

In the GPT-6.2 Astra and Sol launch charts (November 17, 2026), GPT-6.2 Astra leads on Agents' Last Exam v1.1, 77.6% to 72.8%; Claude Fable 5.2 leads on FrontierCode 2.0, 45.6% to 44.2%; Claude Fable 5.2 leads on BenchCAD 2.1 Core, 83.5% to 82.5%. These are OpenAI's figures, not independent measurements.

Which has the larger context window, Claude Fable 5.2 or GPT-6.2 Astra?

GPT-6.2 Astra has the larger context window, 1,050,000 vs 1,000,000 tokens. Both return up to 128,000 tokens per response.

AI Model Brief

The AI models worth knowing, once a week

What shipped, how it scores on independent benchmarks, what it costs per million tokens and what was deprecated, from the data behind our AI model benchmarks.

One email a week. We send a confirmation link first, and every issue has a one-click unsubscribe.

We use your address only to send the brief. See our privacy policy.

See this week’s brief

Production evaluation

Choosing between Claude Fable 5.2 and GPT-6.2 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