Skip to content
A little clarity. A better model.
WORKSPACE
EXPLORE
IN THE ATLAS
More than a leaderboard.

Understand what the numbers actually mean.

Snapshot · Sep 04, 2026
Workspace Overview
Independent. Source-backed.
✳ THE AI MODEL EXPLORER

Find your next model.

Less hype. More perspective. Compare the models shaping what’s possible.

18 models4 benchmark tracksVerified September 4, 2026

The model landscape 18

Find the right balance of capability, cost and control.

18 models · 13 with a DeepSWE resultHigher score is better
AI model comparison sorted by deepSwe, descending
Compare#Score / 100Tokens/ 1M tokens/ 1M tokensAccessSave model
01
74.1
1.05M$10.00$50.00Proprietary
02
73.8
1M$0.75$3.75Proprietary
03
73.7
1M$5.00$25.00Proprietary
04
72.7
1.05M$4.00$20.00Proprietary
05
69.6
—$2.00$12.00Proprietary
06
67.5
1M——Open weights
07
67.4
1M$10.00$50.00Proprietary
08
67.2
—$0.20$1.20Proprietary
01
DeepSWE74.1/100
Context1.05M
Input / 1M$10.00
Output / 1M$50.00
Proprietary
02
DeepSWE73.8/100
Context1M
Input / 1M$0.75
Output / 1M$3.75
Proprietary
03
DeepSWE73.7/100
Context1M
Input / 1M$5.00
Output / 1M$25.00
Proprietary
04
DeepSWE72.7/100
Context1.05M
Input / 1M$4.00
Output / 1M$20.00
Proprietary
05
DeepSWE69.6/100
Context—
Input / 1M$2.00
Output / 1M$12.00
Proprietary
06
DeepSWE67.5/100
Context1M
Input / 1M—
Output / 1M—
Open weights
07
DeepSWE67.4/100
Context1M
Input / 1M$10.00
Output / 1M$50.00
Proprietary
08
DeepSWE67.2/100
Context—
Input / 1M$0.20
Output / 1M$1.20
Proprietary
Showing 8 of 18 models

A benchmark is a starting point, not the whole story. Scores depend on evaluation settings.

Great models. Different strengths.

Select up to three models to see the trade-offs side by side.

Made for curious minds and better decisions.Model Atlas © 2026