Claude Fable 5 vs GPT-6 Astra

Change models
IndexAll categories, out of 100.
74.0 ±2.7
78.0 ±1.9 best
GapIndex points below the top.
−4.0
Top
CoverageIndex weight that has results.
95%
100% best
SpeedOutput tokens per second.
56/s best
34/s
EloLMArena rating and rank.
1492 best
1440
ReleasedMonth of release.
2026-06
2026-09
Input / 1MUS dollars per 1M input tokens.
$10.00
$10.00
Output / 1MUS dollars per 1M output tokens.
$50.00
$50.00
Batch output / 1MBatch price per 1M output tokens.
$25.00
$25.00
1K chats2K tokens in, 500 out, each.
$45.00
$45.00
1K documents20K tokens in, 1K out, each.
$250.00
$250.00
1K agent tasks200K tokens in, 20K out, each.
$3,000.00
$3,000.00
ContextMaximum tokens in one request.
1M
1.05M best
AgenticMulti-step tasks with tools.
71.4
74.6 best
CodingCode writing and repair.
74.7
76.9 best
ReasoningLogic problems and puzzles.
73.6
83.8 best
MultimodalTasks with images and text.
70.1
71.2 best
KnowledgeFacts and expert knowledge.
75.8
78.5 best
MultilingualTasks in many languages.
—
73.6
InstructionTasks with strict prompt rules.
64.1
79.5 best
MathMath problems.
76.2
78.0 best

Benchmarks

59 shared
Only Claude Fable 5: 33Only GPT-6 Astra: 79
Claude Fable 5
GPT-6 Astra
55.5% best5 / 222
54.7%8 / 222
82.3% best27 / 217
80.7%37 / 217
85.9%69 / 217
95.8% best3 / 217
28.6%14 / 214
31.7% best4 / 214
99.7%14 / 200
100.0% best1 / 200
92.6%17 / 199
96.1% best1 / 199
65.4% best4 / 164
61.6%6 / 164
55.5% best18 / 133
53.7%25 / 133
89.2%11 / 131
95.0% best1 / 131
98.5% best1 / 126
97.5%5 / 126
90.4% best4 / 123
89.6%8 / 123
96.4%9 / 122
97.5% best5 / 122
97.6%17 / 116
97.8% best14 / 116
61.0% best6 / 114
56.5%24 / 114
MedScribeHealthcare
88.5% best12 / 109
87.9%16 / 109
MedCodeHealthcare
56.1% best4 / 106
48.5%29 / 106
70.7%9 / 91
75.6% best4 / 91
Chess PuzzlesReasoning
47.0%24 / 90
74.0% best3 / 90
87.0%12 / 90
93.7% best2 / 90
49.5% best9 / 90
39.4%37 / 90