[KongchangAI]
Benchmark

Frontier Code 2.0

A benchmark test used to evaluate the coding capabilities of AI models.

Core Facts

Timeline (last 90 days)

Nov 21

Opus 5.5 leads by about 1 point on Terminal Bench 4.1 and Frontier Code 2.0, while Sol 6.2 has a slight edge on Chartography.

Unverified50%
Nov 21

Google's official comparison table shows that Opus 5.5 leads Neon by about one percentage point on the two coding tests Terminal Bench 4.1 and Frontier Code 2.0.

Unverified50%
Nov 21

On Frontier Code 2.0, Opus led Neon by 1.1 percentage points, and Sol placed third.

Unverified50%
Nov 19

On the Frontier Code 2.0 coding benchmark, Fable 5.2 leads, but all models' scores cluster in the 40s range.

Unverified50%
Nov 19

On Frontier Code 2.0, Gemini 4 Neon scored 46.2%, and Opus 5.5 scored 47.3%.

Partially Verified75%
Nov 13

Anthropic's internal testing shows that Fable 5.2 achieved 49.1% at the highest thinking level on Frontier Code 2.0, up from 32.4% for the previous generation, and scored 69.9% on the Frontier Suite VR.

Unverified50%
Nov 13

On the Frontier Code 2.0 benchmark, Fable 5.2 scored 9.1%, Opus 4.5 scored 45.5, GPT-6 Astra scored 42.8, and Carbon scored 41.1.

Unverified50%
Nov 12

On the Frontier Code 2.0 benchmark, Fable 5.2 at the High setting matches the score of Opus 5.5 at the Max top setting, at about half the latter's per-task cost; at the X High top setting it set a new score above all released models.

Unverified50%
Nov 10

On Frontier code 2.0, Fable 5.2 scored higher at the extra high setting than at the max setting, indicating that a higher reasoning setting does not necessarily bring better performance.

Unverified50%
Nov 10

On the Frontier code benchmark, Fable 5.2 improved from Opus 5's 45.5% to 49.1%; on Frontier SWE v2 it improved by about 7 percentage points.

Unverified50%

12 more timeline events

All Facts (20)

Verified

Claude Fable 5.2 scored 49.1% on the FrontierCode 2.0 benchmark, ranking first.

70%
Verified

Claude Opus 5 approaches Fable-level performance on the FrontierCode 1.1 benchmark while costing only half as much

70%
Verified

Anthropic released two new models, Claude Fable 5.2 and Claude Mythos 5.2.

80%
Partially Verified

On Frontier Code 2.0, Gemini 4 Neon scored 46.2%, and Opus 5.5 scored 47.3%.

75%
Unverified

Opus 5.5 leads by about 1 point on Terminal Bench 4.1 and Frontier Code 2.0, while Sol 6.2 has a slight edge on Chartography.

50%
Unverified

Google's official comparison table shows that Opus 5.5 leads Neon by about one percentage point on the two coding tests Terminal Bench 4.1 and Frontier Code 2.0.

50%
Unverified

On Frontier Code 2.0, Opus led Neon by 1.1 percentage points, and Sol placed third.

50%
Unverified

On the Frontier Code 2.0 coding benchmark, Fable 5.2 leads, but all models' scores cluster in the 40s range.

50%
Unverified

Anthropic's internal testing shows that Fable 5.2 achieved 49.1% at the highest thinking level on Frontier Code 2.0, up from 32.4% for the previous generation, and scored 69.9% on the Frontier Suite VR.

50%
Unverified

On the Frontier Code 2.0 benchmark, Fable 5.2 scored 9.1%, Opus 4.5 scored 45.5, GPT-6 Astra scored 42.8, and Carbon scored 41.1.

50%
Unverified

On the Frontier Code 2.0 benchmark, Fable 5.2 at the High setting matches the score of Opus 5.5 at the Max top setting, at about half the latter's per-task cost; at the X High top setting it set a new score above all released models.

50%
Unverified

On the Frontier code benchmark, Fable 5.2 improved from Opus 5's 45.5% to 49.1%; on Frontier SWE v2 it improved by about 7 percentage points.

50%
Unverified

On Frontier code 2.0, Fable 5.2 scored higher at the extra high setting than at the max setting, indicating that a higher reasoning setting does not necessarily bring better performance.

50%
Unverified

With the X High setting enabled, Fable 5.2 scored the highest among all released models on the early Frontier Code 2.0 test.

50%
Unverified

On the agentic coding Frontier Code 2.0 test, Fable 5.2 scored second highest, behind only Opus 5.5.

50%
Unverified

On the Frontier Code 2.0 coding benchmark, Fable 5.2 scored 49.1%, compared with 32.4% for the previous generation Fable 5.1.

50%
Unverified

Benchmark tests show that Fable 5.2 comprehensively beats the previous flagship Opus 5.5 on major leaderboards such as Frontier Code 2.0 and Frontier Suite (69.9 points).

50%
Unverified

An anonymous model, Obsidian 3, appeared in Cognition's new Frontier Code 2.0 benchmark, scoring 46.8%, surpassing Opus 5.5's 45.5%.

50%
Unverified

Artificial Analysis gave GPT-6.1 Sol a score of 52, only one point below Astra, with benchmark run costs at one-quarter of Astra's and 31% cheaper than 6 Sol.

50%
Unverified

On Frontier Code, Opus 5.5's medium tier (under $1 per task) actually scored higher than its max tier (over $5 per task).

50%

Source Articles