Grade My Model

Explore DeepSWE model performance, pricing, and comparisons in one place.

Leaderboard snapshot: 2026-08-13T16:11:55.708636+00:00

Model selector#

Choose the model and reasoning effort rows to compare.

15 of 61 selected

Show outdated models (16 hidden)Reset to latest

Anthropic

Claude Fable

Select all

xhigh69.9% · $13.41max69.7% · $21.63high68.6% · $9.18medium65.4% · $6.09low59.6% · $3.76

Anthropic

Claude Opus

Select all

max73.6% · $11.84xhigh73.2% · $9.07high72.8% · $6.08medium68.9% · $3.29low58.1% · $1.66

Anthropic

Claude Sonnet

Select all

max53.8% · $26.40xhigh49.7% · $11.89high48.2% · $7.43medium39.8% · $4.08low30.5% · $2.19

DeepSeek

Deepseek V Flash

Select none

max53.3% · $0.10

DeepSeek

Deepseek V Pro

Select none

max62.8% · $0.24

Google

Gemini 3 Flash

Select all

medium65.5% · $2.03high65.3% · $2.18low53.8% · $1.83

Zhipu

Glm

Select all

max43.8% · $3.92high36.3% · $2.84

Luna

Gpt 5 Luna

Select all

max67.2% · $3.03xhigh56.9% · $1.54high44.2% · $0.78medium11.3% · $0.22low1.5% · $0.07

OpenAI

Gpt 5 Sol

Select all

max72.7% · $8.39xhigh70.7% · $4.70high69.4% · $3.47medium61.1% · $1.86low45.4% · $1.07

OpenAI

Gpt 5 Terra

Select all

max69.6% · $4.95xhigh60.2% · $2.13high53.8% · $1.13medium35.1% · $0.58low24.1% · $0.43

xAI

Grok 4

Select all

medium67.5% · $3.45xhigh66.7% · $5.50high65.2% · $4.38low41.6% · $1.04

Moonshot

Kimi K

Select none

max68.5% · $4.65

Moonshot

Kimi K Code

Select none

default30.5% · $2.82

Meta

Muse Spark

Select none

xhigh54.9% · $3.70

Alibaba

Qwen Max

Select none

xhigh57.5% · $3.73

Leaderboard table#

FilledUncertainEmptyShow speed (output tokens + agent steps)

15 selected models · sort by performance, price, or speed when shown

Model	Performance↓	Price↕lower is cheaper	Savings	Reference
claude-opus-5 max Anthropic	73.6% ±3.9	$11.84 / task $0.16/ point	60.3% cheaper reference	gpt-5-6-sol Tier: xhigh
gpt-5-6-sol max OpenAI	72.7% ±2.8	$8.39 / task $0.12/ point	43.9% cheaper reference	gpt-5-6-sol Tier: xhigh
claude-fable-5 xhigh Anthropic	69.9% ±3.2	$13.41 / task $0.19/ point	84.9% cheaper reference	gemini-3-7-flash Tier: medium
gpt-5-6-terra max OpenAI	69.6% ±2.6	$4.95 / task $0.07/ point	59.1% cheaper reference	gemini-3-7-flash Tier: medium
kimi-k3 max Moonshot	68.5% ±4.5	$4.65 / task $0.07/ point	56.5% cheaper reference	gemini-3-7-flash Tier: medium
grok-4-6 medium xAI	67.5% ±2.3	$3.45 / task $0.05/ point	41.3% cheaper reference	gemini-3-7-flash Tier: medium
gpt-5-6-luna max Luna	67.2% ±4.0	$3.03 / task $0.05/ point	33.1% cheaper reference	gemini-3-7-flash Tier: medium
gemini-3-7-flash medium Google	65.5% ±3.1	$2.03 / task $0.03/ point	—	No cheaper match Similar-score reference unavailable
deepseek-v4-pro max DeepSeek	62.8% ±6.3	$0.24 / task $0.0038/ point	—	No cheaper match Similar-score reference unavailable
qwen3-8-max xhigh Alibaba	57.5% ±2.7	$3.73 / task $0.06/ point	93.5% cheaper reference	deepseek-v4-pro Tier: max
muse-spark-1-2 xhigh Meta	54.9% ±2.1	$3.70 / task $0.07/ point	97.3% cheaper reference	deepseek-v4-flash Tier: max
claude-sonnet-5 max Anthropic	53.8% ±4.2	$26.40 / task $0.49/ point	99.6% cheaper reference	deepseek-v4-flash Tier: max
deepseek-v4-flash max DeepSeek	53.3% ±3.6	$0.10 / task $0.0019/ point	—	No cheaper match Similar-score reference unavailable
glm-5-2 max Zhipu	43.8% ±1.7	$3.92 / task $0.09/ point	97.4% cheaper reference	deepseek-v4-flash Tier: max
kimi-k2-7-code default Moonshot	30.5% ±0.5	$2.82 / task $0.09/ point	96.4% cheaper reference	deepseek-v4-flash Tier: max