Cookies on AIGROW

We use necessary cookies to run the site. With your permission, we also use cookies for analytics, session recordings and marketing attribution.

Cookie Policy

AI leaderboard · Math problems

AI math benchmark: 92 models ranked on measured results

FrontierMath Tiers 1-3: hundreds of problems written by experts, from advanced undergraduate level to early research, each with an answer a program can check. Epoch AI runs it and keeps all but a few problems private.

Run Claude Fable 5.2 up to 60% below list price. $25 a month brings $110 of API credit and cancels any time; $25 once buys $100. Results from tests Epoch AI runs, data of 12 November 2026, under CC BY 4.0. Prices from each maker’s pricing page, read on 12 November 2026.

Top score
95.1%Claude Fable 5.2 from Anthropic
Models ranked
92
Level with the leader
8
On sale with AIGROW credit
9
Data of
12 November 2026

Level: its interval overlaps the leader’s.

On FrontierMath Tiers 1-3, run by Epoch AI, Claude Fable 5.2 from Anthropic leads 92 models with 95.1% of the problems solved (data of 12 November 2026). 8 other models are statistically level with it. AIGROW API credit sells it at $3.20 / $16.00 instead of $8.00 / $40.00 per million input / output tokens.

Run models from this ranking below list price

AIGROW API credit runs these models through one key, billed per token at the rates below. $25 a month brings $110 of API credit and cancels any time; $25 once buys $100.

  1. No. 1Claude Fable 5.2AnthropicList price $8.00 / $40.00$3.20 / $16.00with AIGROW API creditRun Claude Fable 5.2 up to 60% below list priceCertified genuine by AIGROW on 11 November 2026
  2. No. 2GPT-6.1 SolOpenAIList price $2.00 / $10.00$0.50 / $2.50with AIGROW API creditRun GPT-6.1 Sol up to 75% below list priceCertified genuine by AIGROW on 11 November 2026
  3. No. 3GPT-6 AstraOpenAIList price $10.00 / $50.00$2.50 / $12.50with AIGROW API creditRun GPT-6 Astra up to 75% below list priceCertified genuine by AIGROW on 11 November 2026

US dollars per million input / output tokens. Certified genuine by AIGROW means the model passed AIGROW’s own check against the same model at its maker on 11 November 2026. No model maker reviewed or endorsed it.

AI math benchmark

It checks final answers, not proofs or explanations.

Results of 92 AI models on FrontierMath Tiers 1-3, with their 95% intervals and list prices
#ModelScoreSetting and runPer 1M tokens, in / out
1Claude Fable 5.2AnthropicRun Claude Fable 5.2 up to 60% below list priceCertified genuine by AIGROW on 11 November 202695.1%95% interval 92.6% to 97.6%max, run 6 Nov 2026List price $8.00 / $40.00$3.20 / $16.00with AIGROW API credit
2GPT-6.1 SolOpenAIRun GPT-6.1 Sol up to 75% below list priceCertified genuine by AIGROW on 11 November 202693.7%95% interval 90.9% to 96.5%Its interval overlaps 1 model above itmax, run 29 Sept 2026List price $2.00 / $10.00$0.50 / $2.50with AIGROW API credit
3GPT-6 AstraOpenAIRun GPT-6 Astra up to 75% below list priceCertified genuine by AIGROW on 11 November 202693.7%95% interval 90.9% to 96.5%Its interval overlaps 2 models above itmax, run 30 Aug 2026List price $10.00 / $50.00$2.50 / $12.50with AIGROW API credit
4Claude Opus 5.5AnthropicRun Claude Opus 5.5 up to 60% below list priceCertified genuine by AIGROW on 11 November 202691.2%95% interval 87.9% to 94.5%Its interval overlaps 3 models above itmax, run 22 Sept 2026List price $4.00 / $20.00$1.60 / $8.00with AIGROW API credit
5GPT-6.1 TerraOpenAIRun GPT-6.1 Terra up to 75% below list priceCertified genuine by AIGROW on 11 November 202690.9%95% interval 87.5% to 94.2%Its interval overlaps 4 models above itmax, run 26 Oct 2026List price $1.00 / $5.00$0.25 / $1.25with AIGROW API credit
6Gemini 4 CarbonGoogle DeepMind90.5%95% interval 87.1% to 93.9%Its interval overlaps 5 models above ithigh, run 2 Nov 2026$2.00 / $10.00
7Claude Fable 5.1AnthropicRun Claude Fable 5.1 up to 60% below list priceCertified genuine by AIGROW on 11 November 202690.2%95% interval 86.7% to 93.6%Its interval overlaps 6 models above itmax, run 1 Sept 2026List price $10.00 / $50.00$4.00 / $20.00with AIGROW API credit
8GPT-6 SolOpenAI89.8%95% interval 86.3% to 93.3%Its interval overlaps 7 models above itmax, run 22 Sept 2026$2.00 / $10.00
9GPT-5.6 SolOpenAI89.1%95% interval 85.5% to 92.7%Its interval overlaps 8 models above itmax, run 9 Jul 2026Not updated in over 120 days$4.00 / $20.00
10Claude Sonnet 5.5AnthropicRun Claude Sonnet 5.5 up to 60% below list priceCertified genuine by AIGROW on 11 November 202688.8%95% interval 85.1% to 92.4%Its interval overlaps 8 models above itmax, run 29 Sept 2026List price $2.00 / $10.00$0.80 / $4.00with AIGROW API credit
11GPT-5.5 ProOpenAI87.7%95% interval 83.9% to 91.5%Its interval overlaps 9 models above itxhigh, run 12 Jun 2026Not updated in over 120 daysNo list price
12Claude Fable 5Anthropic87.0%95% interval 83.1% to 90.9%Its interval overlaps 10 models above itmax, run 9 Jun 2026Not updated in over 120 daysNo list price
13Gemini 4 ArgonGoogle DeepMindRun Gemini 4 Argon up to 50% below list priceCertified genuine by AIGROW on 11 November 202686.0%95% interval 81.9% to 90.0%Its interval overlaps 9 models above ithigh, run 13 Oct 2026List price $2.00 / $10.00$1.00 / $5.00with AIGROW API credit
14GPT-5.6 TerraOpenAI86.0%95% interval 81.9% to 90.0%Its interval overlaps 10 models above itmax, run 9 Jul 2026Not updated in over 120 days$2.00 / $12.00
15Claude Opus 5Anthropic85.6%95% interval 81.5% to 89.7%Its interval overlaps 11 models above itmax, run 24 Jul 2026$5.00 / $25.00
16GPT-5.5OpenAI85.3%95% interval 81.1% to 89.4%Its interval overlaps 12 models above itxhigh, run 11 Jun 2026Not updated in over 120 days$5.00 / $30.00
17GPT-5.4 ProOpenAI82.5%95% interval 78.0% to 86.9%Its interval overlaps 10 models above itxhigh, run 13 Jun 2026Not updated in over 120 daysNo list price
18Muse WonderMeta AI82.1%95% interval 77.7% to 86.6%Its interval overlaps 10 models above itmax, run 12 Oct 2026$4.00 / $20.00
19GPT-5.6 LunaOpenAI82.1%95% interval 77.7% to 86.6%Its interval overlaps 11 models above itmax, run 9 Jul 2026Not updated in over 120 days$0.20 / $1.20
20Claude Opus 4.8Anthropic80.0%95% interval 75.4% to 84.6%Its interval overlaps 9 models above itmax, run 10 Jun 2026Not updated in over 120 daysNo list price
21GPT-6 LunaOpenAI79.0%95% interval 74.2% to 83.7%Its interval overlaps 9 models above itmax, run 22 Sept 2026$0.10 / $0.50
22GPT-5.4OpenAI78.6%95% interval 73.8% to 83.4%Its interval overlaps 10 models above itxhigh, run 11 Jun 2026Not updated in over 120 daysNo list price
23Kimi K3.1 ProMoonshot78.3%95% interval 73.4% to 83.1%Its interval overlaps 10 models above itmax, run 6 Nov 2026$2.00 / $10.00
24Claude Haiku 5.5Anthropic75.1%95% interval 70.1% to 80.1%Its interval overlaps 7 models above itmax, run 7 Oct 2026$0.10 / $0.50
25Qwen3.8 MaxAlibaba74.7%95% interval 69.7% to 79.8%Its interval overlaps 8 models above itxhigh, run 4 Aug 2026$2.00 / $6.00

How to read this ranking

What the scores measure, what they leave out, and where each number comes from.

What it measures
FrontierMath Tiers 1-3: hundreds of problems written by experts, from advanced undergraduate level to early research, each with an answer a program can check. Epoch AI runs it and keeps all but a few problems private.
What it leaves out
It checks final answers, not proofs or explanations.
How to read the margin
Each result carries a 95% interval, 1.96 standard errors either side as Epoch AI publishes them; an average combines the errors of its tests. Where two intervals overlap, the tests cannot say which model is better. Rows are still numbered by score, and a row says how many models above it overlap.
Settings
Under each result is the reasoning setting Epoch AI ran the model at, as it names it, and the day of the run. Where a model ran at several settings, the best result counts.
Who funds it
Epoch AI developed FrontierMath with funding from OpenAI, which has exclusive access to a subset of the problems.
Near the ceiling
The top score, 95.1%, has reached the 95.0% that Epoch AI treats as the most a model can score on FrontierMath Tiers 1-3, given the test’s known error rate. Near the ceiling a test tells the leading models apart less well.
Not updated in over 120 days
A result run more than 120 days before the newest data on these pages carries this mark, so an older result can be weighed as one.
Prices
The list price per million tokens from the maker’s own pricing page, read on 12 November 2026. It is a price per token, not per answer: a higher reasoning setting spends more tokens on the same question. Where AIGROW API credit sells a model, the list price is struck through and the price through the credit sits under it.

Source: Epoch AI, FrontierMath Tiers 1-3 (data of 12 November 2026), licensed under CC BY 4.0. We change no score: we add the prices and the links to our model pages.

What each of them costs to run

The price list works three everyday jobs on every model, and our own benchmark runs them on real business tasks to see which pass.

See the pricesSee the benchmark

Data of 12 November 2026