Verified

SWE-bench · 50 models

Rank	Model	Score
1	Sonar Foundation Agent + Claude 4.5 Opus	79.2
1	live-SWE-agent + Claude 4.5 Opus medium (20251101)	79.2
3	TRAE + Doubao-Seed-Code	78.8
4	live-SWE-agent + Gemini 3 Pro Preview (2025-11-18)	77.4
5	Atlassian Rovo Dev (2025-09-02)	76.8
5	Claude 4.5 Opus (high)	76.8
5	EPAM AI/Run Developer Agent v20250719 + Claude 4 Sonnet	76.8
8	ACoder	76.4
9	Gemini 3 Flash (high)	75.8
9	MiniMax M2.5 (high)	75.8
11	Claude 4.6 Opus	75.6
11	Warp	75.6
13	TRAE + Claude Sonnet 4 + Opus 4 + Sonnet 3.7 + Gemini 2.5 Pro	75.2
14	Harness AI	74.8
14	Sonar Foundation Agent + Claude 4.5 Sonnet	74.8
16	JoyCode + Claude 4 Sonnet + GPT-4.1	74.6
16	Lingxi-v1.5_claude-4-sonnet-20250514	74.6
18	Claude 4.5 Opus (20251101) (medium)	74.4
18	Prometheus-v1.2.1 + GPT-5	74.4
18	Refact.ai Agent + Claude 4 Sonnet + o4-mini	74.4
21	Gemini 3 Pro Preview (2025-11-18)	74.2
22	Salesforce AI Research SAGE (OpenHands)	73.8
23	Tools + Claude 4 Opus (2025-05-22)	73.2
24	Salesforce AI Research SAGE (bash-only)	73
25	GLM 5 (high)	72.8
25	GPT 5.2 (high)	72.8
25	GPT 5.2 Codex	72.8
28	Tools + Claude 4 Sonnet (2025-05-22)	72.4
29	GPT 5.2 (2025-12-11) (high)	71.8
29	OpenHands + GPT-5	71.8
31	Claude 4.5 Sonnet (high)	71.4
32	Bloop	71.2
32	Lingxi v1.5 x Kimi K2	71.2
32	Prometheus-v1.2 + GPT-5	71.2
32	Qodo Command	71.2
36	Warp	71
37	Kimi K2.5 (high)	70.8
37	Moatless Tools + Claude 4 Sonnet	70.8
39	Claude 4.5 Sonnet (20250929)	70.6
39	TRAE	70.6
41	Augment Agent v1	70.4
41	OpenHands + Claude 4 Sonnet	70.4
41	Refact.ai Agent	70.4
44	devlo	70.2
45	DeepSeek V3.2 (high)	70
45	Zencoder (2025-04-30)	70
47	Gemini 3 Pro	69.6
47	OpenHands + Qwen3-Coder-480B-A35B-Instruct	69.6
49	GPT 5.2 (2025-12-11)	69
50	GLM-4.6	68.2