Back to AI Coding

AI Model Ranking

AI Coding Model Rankings

Top 50 from Artificial Analysis.

Snapshot: 2026-11-11

Source Artificial AnalysisMethodology

Metrics follow a snapshot of the Artificial Analysis public LLM leaderboard: Intelligence Index, cost per task, and speed reflect benchmark performance at the snapshot date. Provider pricing, availability, and rankings change frequently, and a high benchmark score does not guarantee the best result for a specific codebase or workflow.

Read the full methodology guide

AA Index top 50

65
62
62
60
59
58
57
57
56
55
55
55
55
54
54
52
52
52
51
50
50
50
50
48
46
46
45
44
43
42
42
41
38
36
36
35
35
35
35
35
35
34
34
33
32
31
31
31
31
31

Top 50 models

01

Current reasoning model

Anthropic - Claude Fable - 5.2 (max)

Anthropic Claude Fable 5.2 (max) leads the 2026-11-11 snapshot with Intelligence Index 65, 1M context, $26.54 cost per task, 101 tok/s median output, and 233.95s to first chunk. First chunk takes 233.95s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort Anthropic work on hard refactors and architecture decisions, at $26.54 per task with 233.95s to first chunk.

Anthropic

Context
1M
AA Index
65
Cost per task
$26.54
Speed
101 tok/s
First chunk
233.95s
Total response
238.90s
02

Current reasoning model

Anthropic - Claude Fable - 5.2 (xhigh)

Anthropic Claude Fable 5.2 (xhigh) ranks #2 in the 2026-11-11 snapshot with Intelligence Index 62, 1M context, $14.83 cost per task, 103 tok/s median output, and 69.26s to first chunk. It is one of 2 models at Intelligence Index 62. First chunk takes 69.26s, so it fits batch or async work better than interactive loops.

Best for: Very high-effort Anthropic work where a near-top index is worth a longer wait, at $14.83 per task with 69.26s to first chunk.

Anthropic

Context
1M
AA Index
62
Cost per task
$14.83
Speed
103 tok/s
First chunk
69.26s
Total response
74.10s
03

Current reasoning model

Anthropic - Claude Fable - 5.2 (high)

Anthropic Claude Fable 5.2 (high) ranks #3 in the 2026-11-11 snapshot with Intelligence Index 62, 1M context, $9.08 cost per task, 94 tok/s median output, and 30.40s to first chunk. It is one of 2 models at Intelligence Index 62.

Best for: High-effort Anthropic coding and multi-step agent runs, at $9.08 per task with 30.40s to first chunk.

Anthropic

Context
1M
AA Index
62
Cost per task
$9.08
Speed
94 tok/s
First chunk
30.40s
Total response
35.70s
04

Current reasoning model

Google - Gemini - 4 Carbon (high)

Google Gemini 4 Carbon (high) ranks #4 in the 2026-11-11 snapshot with Intelligence Index 60, 1M context, $7.76 cost per task, 80 tok/s median output, and 67.08s to first chunk. First chunk takes 67.08s, so it fits batch or async work better than interactive loops.

Best for: High-effort Google coding and multi-step agent runs, at $7.76 per task with 67.08s to first chunk.

Google

Context
1M
AA Index
60
Cost per task
$7.76
Speed
80 tok/s
First chunk
67.08s
Total response
73.33s
05

Current reasoning model

Anthropic - Claude Opus - 5.5 (max)

Anthropic Claude Opus 5.5 (max) ranks #5 in the 2026-11-11 snapshot with Intelligence Index 59, 1M context, $10.04 cost per task, 115 tok/s median output, and 356.37s to first chunk. First chunk takes 356.37s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort Anthropic work on hard refactors and architecture decisions, at $10.04 per task with 356.37s to first chunk.

Anthropic

Context
1M
AA Index
59
Cost per task
$10.04
Speed
115 tok/s
First chunk
356.37s
Total response
360.70s
06

Current reasoning model

Google - Gemini - 4 Carbon (medium)

Google Gemini 4 Carbon (medium) ranks #6 in the 2026-11-11 snapshot with Intelligence Index 58, 1M context, $6.59 cost per task, 74 tok/s median output, and 23.23s to first chunk.

Best for: Balanced Google coding with a moderate task budget, at $6.59 per task with 23.23s to first chunk.

Google

Context
1M
AA Index
58
Cost per task
$6.59
Speed
74 tok/s
First chunk
23.23s
Total response
30.02s
07

Current reasoning model

Anthropic - Claude Opus - 5.5 (xhigh)

Anthropic Claude Opus 5.5 (xhigh) ranks #7 in the 2026-11-11 snapshot with Intelligence Index 57, 1M context, $6.05 cost per task, 110 tok/s median output, and 80.27s to first chunk. It is one of 2 models at Intelligence Index 57. First chunk takes 80.27s, so it fits batch or async work better than interactive loops.

Best for: Very high-effort Anthropic work where a near-top index is worth a longer wait, at $6.05 per task with 80.27s to first chunk.

Anthropic

Context
1M
AA Index
57
Cost per task
$6.05
Speed
110 tok/s
First chunk
80.27s
Total response
84.83s
08

Current reasoning model

OpenAI - GPT Astra - 6 (max)

OpenAI GPT Astra 6 (max) ranks #8 in the 2026-11-11 snapshot with Intelligence Index 57, 1M context, $6.24 cost per task, 38 tok/s median output, and 244.09s to first chunk. It is one of 2 models at Intelligence Index 57. First chunk takes 244.09s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort OpenAI work on hard refactors and architecture decisions, at $6.24 per task with 244.09s to first chunk.

OpenAI

Context
1M
AA Index
57
Cost per task
$6.24
Speed
38 tok/s
First chunk
244.09s
Total response
257.31s
09

Current reasoning model

Anthropic - Claude Fable - 5.2 (medium)

Anthropic Claude Fable 5.2 (medium) ranks #9 in the 2026-11-11 snapshot with Intelligence Index 56, 1M context, $3.80 cost per task, 93 tok/s median output, and 12.76s to first chunk.

Best for: Balanced Anthropic coding with a moderate task budget, at $3.80 per task with 12.76s to first chunk.

Anthropic

Context
1M
AA Index
56
Cost per task
$3.80
Speed
93 tok/s
First chunk
12.76s
Total response
18.16s
10

Current reasoning model

OpenAI - GPT Astra - 6 (xhigh)

OpenAI GPT Astra 6 (xhigh) ranks #10 in the 2026-11-11 snapshot with Intelligence Index 55, 1M context, $3.66 cost per task, 39 tok/s median output, and 110.88s to first chunk. It is one of 4 models at Intelligence Index 55. First chunk takes 110.88s, so it fits batch or async work better than interactive loops.

Best for: Very high-effort OpenAI work where a near-top index is worth a longer wait, at $3.66 per task with 110.88s to first chunk.

OpenAI

Context
1M
AA Index
55
Cost per task
$3.66
Speed
39 tok/s
First chunk
110.88s
Total response
123.56s
11

Current reasoning model

OpenAI - GPT Astra - 6 (high)

OpenAI GPT Astra 6 (high) ranks #11 in the 2026-11-11 snapshot with Intelligence Index 55, 1M context, $3.24 cost per task, 40 tok/s median output, and 65.27s to first chunk. It is one of 4 models at Intelligence Index 55. First chunk takes 65.27s, so it fits batch or async work better than interactive loops.

Best for: High-effort OpenAI coding and multi-step agent runs, at $3.24 per task with 65.27s to first chunk.

OpenAI

Context
1M
AA Index
55
Cost per task
$3.24
Speed
40 tok/s
First chunk
65.27s
Total response
77.84s
12

Current reasoning model

OpenAI - GPT Sol - 6.1 (max)

OpenAI GPT Sol 6.1 (max) ranks #12 in the 2026-11-11 snapshot with Intelligence Index 55, 1M context, $1.27 cost per task, 56 tok/s median output, and 153.34s to first chunk. It is one of 4 models at Intelligence Index 55. First chunk takes 153.34s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort OpenAI work on hard refactors and architecture decisions, at $1.27 per task with 153.34s to first chunk.

OpenAI

Context
1M
AA Index
55
Cost per task
$1.27
Speed
56 tok/s
First chunk
153.34s
Total response
162.27s
13

Current reasoning model

Anthropic - Claude Sonnet - 5.5 (max)

Anthropic Claude Sonnet 5.5 (max) ranks #13 in the 2026-11-11 snapshot with Intelligence Index 55, 1M context, $10.02 cost per task, 145 tok/s median output, and 255.96s to first chunk. It is one of 4 models at Intelligence Index 55. First chunk takes 255.96s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort Anthropic work on hard refactors and architecture decisions, at $10.02 per task with 255.96s to first chunk.

Anthropic

Context
1M
AA Index
55
Cost per task
$10.02
Speed
145 tok/s
First chunk
255.96s
Total response
259.40s
14

Current reasoning model

Google - Gemini - 4 Carbon (low)

Google Gemini 4 Carbon (low) ranks #14 in the 2026-11-11 snapshot with Intelligence Index 54, 1M context, $7.19 cost per task, 63 tok/s median output, and 12.90s to first chunk. It is one of 2 models at Intelligence Index 54.

Best for: Light Google passes and high-volume runs, at $7.19 per task with 12.90s to first chunk.

Google

Context
1M
AA Index
54
Cost per task
$7.19
Speed
63 tok/s
First chunk
12.90s
Total response
20.78s
15

Current reasoning model

Anthropic - Claude Fable - 5.2 (low)

Anthropic Claude Fable 5.2 (low) ranks #15 in the 2026-11-11 snapshot with Intelligence Index 54, 1M context, $3.10 cost per task, 94 tok/s median output, and 8.59s to first chunk. It is one of 2 models at Intelligence Index 54.

Best for: Light Anthropic passes and high-volume runs, at $3.10 per task with 8.59s to first chunk.

Anthropic

Context
1M
AA Index
54
Cost per task
$3.10
Speed
94 tok/s
First chunk
8.59s
Total response
13.89s
16

Current reasoning model

OpenAI - GPT Sol - 6.1 (xhigh)

OpenAI GPT Sol 6.1 (xhigh) ranks #16 in the 2026-11-11 snapshot with Intelligence Index 52, 1M context, $0.76 cost per task, 56 tok/s median output, and 63.35s to first chunk. It is one of 3 models at Intelligence Index 52. First chunk takes 63.35s, so it fits batch or async work better than interactive loops.

Best for: Very high-effort OpenAI work where a near-top index is worth a longer wait, at $0.76 per task with 63.35s to first chunk.

OpenAI

Context
1M
AA Index
52
Cost per task
$0.76
Speed
56 tok/s
First chunk
63.35s
Total response
72.32s
17

Current reasoning model

Anthropic - Claude Opus - 5.5 (high)

Anthropic Claude Opus 5.5 (high) ranks #17 in the 2026-11-11 snapshot with Intelligence Index 52, 1M context, $3.46 cost per task, 106 tok/s median output, and 24.32s to first chunk. It is one of 3 models at Intelligence Index 52.

Best for: High-effort Anthropic coding and multi-step agent runs, at $3.46 per task with 24.32s to first chunk.

Anthropic

Context
1M
AA Index
52
Cost per task
$3.46
Speed
106 tok/s
First chunk
24.32s
Total response
29.05s
18

Current reasoning model

OpenAI - GPT Astra - 6 (medium)

OpenAI GPT Astra 6 (medium) ranks #18 in the 2026-11-11 snapshot with Intelligence Index 52, 1M context, $2.73 cost per task, 43 tok/s median output, and 5.87s to first chunk. It is one of 3 models at Intelligence Index 52.

Best for: Balanced OpenAI coding with a moderate task budget, at $2.73 per task with 5.87s to first chunk.

OpenAI

Context
1M
AA Index
52
Cost per task
$2.73
Speed
43 tok/s
First chunk
5.87s
Total response
17.53s
19

Current reasoning model

OpenAI - GPT Terra - 6.1 (max)

OpenAI GPT Terra 6.1 (max) ranks #19 in the 2026-11-11 snapshot with Intelligence Index 51, 1M context, $0.96 cost per task, 74 tok/s median output, and 155.82s to first chunk. First chunk takes 155.82s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort OpenAI work on hard refactors and architecture decisions, at $0.96 per task with 155.82s to first chunk.

OpenAI

Context
1M
AA Index
51
Cost per task
$0.96
Speed
74 tok/s
First chunk
155.82s
Total response
162.58s
20

Current reasoning model

Google - Gemini - 4 Argon (high)

Google Gemini 4 Argon (high) ranks #20 in the 2026-11-11 snapshot with Intelligence Index 50, 1M context, $9.95 cost per task, 67 tok/s median output, and 40.12s to first chunk. It is one of 4 models at Intelligence Index 50.

Best for: High-effort Google coding and multi-step agent runs, at $9.95 per task with 40.12s to first chunk.

Google

Context
1M
AA Index
50
Cost per task
$9.95
Speed
67 tok/s
First chunk
40.12s
Total response
47.60s
21

Current reasoning model

OpenAI - GPT Sol - 6.1 (high)

OpenAI GPT Sol 6.1 (high) ranks #21 in the 2026-11-11 snapshot with Intelligence Index 50, 1M context, $0.59 cost per task, 57 tok/s median output, and 34.35s to first chunk. It is one of 4 models at Intelligence Index 50.

Best for: High-effort OpenAI coding and multi-step agent runs, at $0.59 per task with 34.35s to first chunk.

OpenAI

Context
1M
AA Index
50
Cost per task
$0.59
Speed
57 tok/s
First chunk
34.35s
Total response
43.19s
22

Current model; partial public metrics

Meta - Muse Wonder Full - Max

Meta Muse Wonder Full Max ranks #22 in the 2026-11-11 snapshot with Intelligence Index 50 and 1M context. Cost per task, output speed, and latency are not published yet. It is one of 4 models at Intelligence Index 50.

Best for: Maximum-effort Meta work on hard refactors and architecture decisions; cost, speed, and latency are not published yet.

Meta

Context
1M
AA Index
50
Cost per task
—
Speed
— tok/s
First chunk
—
Total response
—
23

Current reasoning model

Anthropic - Claude Sonnet - 5.5 (xhigh)

Anthropic Claude Sonnet 5.5 (xhigh) ranks #23 in the 2026-11-11 snapshot with Intelligence Index 50, 1M context, $4.09 cost per task, 139 tok/s median output, and 27.62s to first chunk. It is one of 4 models at Intelligence Index 50.

Best for: Very high-effort Anthropic work where a near-top index is worth a longer wait, at $4.09 per task with 27.62s to first chunk.

Anthropic

Context
1M
AA Index
50
Cost per task
$4.09
Speed
139 tok/s
First chunk
27.62s
Total response
31.20s
24

Current reasoning model

Anthropic - Claude Opus - 5.5 (medium)

Anthropic Claude Opus 5.5 (medium) ranks #24 in the 2026-11-11 snapshot with Intelligence Index 48, 1M context, $2.45 cost per task, 103 tok/s median output, and 14.51s to first chunk.

Best for: Balanced Anthropic coding with a moderate task budget, at $2.45 per task with 14.51s to first chunk.

Anthropic

Context
1M
AA Index
48
Cost per task
$2.45
Speed
103 tok/s
First chunk
14.51s
Total response
19.34s
25

Current reasoning model

Kimi - Kimi - K3.1 Pro

Kimi Kimi K3.1 Pro ranks #25 in the 2026-11-11 snapshot with Intelligence Index 46, 1M context, $3.68 cost per task, 32 tok/s median output, and 169.38s to first chunk. It is one of 2 models at Intelligence Index 46. First chunk takes 169.38s, so it fits batch or async work better than interactive loops.

Best for: Kimi's flagship Kimi model for demanding coding work, at $3.68 per task with 169.38s to first chunk.

Kimi

Context
1M
AA Index
46
Cost per task
$3.68
Speed
32 tok/s
First chunk
169.38s
Total response
184.83s
26

Current reasoning model

OpenAI - GPT Sol - 6.1 (medium)

OpenAI GPT Sol 6.1 (medium) ranks #26 in the 2026-11-11 snapshot with Intelligence Index 46, 1M context, $0.36 cost per task, 58 tok/s median output, and 4.43s to first chunk. It is one of 2 models at Intelligence Index 46. First chunk arrives in 4.43s, so it suits interactive loops.

Best for: Balanced OpenAI coding with a moderate task budget, at $0.36 per task with 4.43s to first chunk.

OpenAI

Context
1M
AA Index
46
Cost per task
$0.36
Speed
58 tok/s
First chunk
4.43s
Total response
13.02s
27

Current reasoning model

OpenAI - GPT Terra - 6.1 (xhigh)

OpenAI GPT Terra 6.1 (xhigh) ranks #27 in the 2026-11-11 snapshot with Intelligence Index 45, 1M context, $0.40 cost per task, 72 tok/s median output, and 28.18s to first chunk.

Best for: Very high-effort OpenAI work where a near-top index is worth a longer wait, at $0.40 per task with 28.18s to first chunk.

OpenAI

Context
1M
AA Index
45
Cost per task
$0.40
Speed
72 tok/s
First chunk
28.18s
Total response
35.09s
28

Current reasoning model

Meta - Muse Wonder - Max

Meta Muse Wonder Max ranks #28 in the 2026-11-11 snapshot with Intelligence Index 44, 1.05M context, $6.95 cost per task, 121 tok/s median output, and 22.40s to first chunk.

Best for: Maximum-effort Meta work on hard refactors and architecture decisions, at $6.95 per task with 22.40s to first chunk.

Meta

Context
1.05M
AA Index
44
Cost per task
$6.95
Speed
121 tok/s
First chunk
22.40s
Total response
26.52s
29

Current reasoning model

OpenAI - GPT Astra - 6 (low)

OpenAI GPT Astra 6 (low) ranks #29 in the 2026-11-11 snapshot with Intelligence Index 43, 1M context, $1.42 cost per task, 40 tok/s median output, and 3.57s to first chunk. First chunk arrives in 3.57s, so it suits interactive loops.

Best for: Light OpenAI passes and high-volume runs, at $1.42 per task with 3.57s to first chunk.

OpenAI

Context
1M
AA Index
43
Cost per task
$1.42
Speed
40 tok/s
First chunk
3.57s
Total response
16.02s
30

Current reasoning model

OpenAI - GPT Terra - 6.1 (high)

OpenAI GPT Terra 6.1 (high) ranks #30 in the 2026-11-11 snapshot with Intelligence Index 42, 1M context, $0.26 cost per task, 70 tok/s median output, and 10.48s to first chunk. It is one of 2 models at Intelligence Index 42.

Best for: High-effort OpenAI coding and multi-step agent runs, at $0.26 per task with 10.48s to first chunk.

OpenAI

Context
1M
AA Index
42
Cost per task
$0.26
Speed
70 tok/s
First chunk
10.48s
Total response
17.60s
31

Current reasoning model

Anthropic - Claude Sonnet - 5.5 (high)

Anthropic Claude Sonnet 5.5 (high) ranks #31 in the 2026-11-11 snapshot with Intelligence Index 42, 1M context, $1.64 cost per task, 141 tok/s median output, and 10.18s to first chunk. It is one of 2 models at Intelligence Index 42.

Best for: High-effort Anthropic coding and multi-step agent runs, at $1.64 per task with 10.18s to first chunk.

Anthropic

Context
1M
AA Index
42
Cost per task
$1.64
Speed
141 tok/s
First chunk
10.18s
Total response
13.73s
32

Current reasoning model

Meta - Muse Wonder - High

Meta Muse Wonder High ranks #32 in the 2026-11-11 snapshot with Intelligence Index 41, 1.05M context, $4.56 cost per task, 122 tok/s median output, and 9.64s to first chunk.

Best for: High-effort Meta coding and multi-step agent runs, at $4.56 per task with 9.64s to first chunk.

Meta

Context
1.05M
AA Index
41
Cost per task
$4.56
Speed
122 tok/s
First chunk
9.64s
Total response
13.75s
33

Current reasoning model

OpenAI - GPT Terra - 6.1 (medium)

OpenAI GPT Terra 6.1 (medium) ranks #33 in the 2026-11-11 snapshot with Intelligence Index 38, 1M context, $0.17 cost per task, 70 tok/s median output, and 4.16s to first chunk. First chunk arrives in 4.16s, so it suits interactive loops.

Best for: Balanced OpenAI coding with a moderate task budget, at $0.17 per task with 4.16s to first chunk.

OpenAI

Context
1M
AA Index
38
Cost per task
$0.17
Speed
70 tok/s
First chunk
4.16s
Total response
11.35s
34

Current reasoning model

Anthropic - Claude Haiku - 5.5 (max)

Anthropic Claude Haiku 5.5 (max) ranks #34 in the 2026-11-11 snapshot with Intelligence Index 36, 1M context, $0.41 cost per task, 255 tok/s median output, and 186.14s to first chunk. It is one of 2 models at Intelligence Index 36. First chunk takes 186.14s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort Anthropic work on hard refactors and architecture decisions, at $0.41 per task with 186.14s to first chunk.

Anthropic

Context
1M
AA Index
36
Cost per task
$0.41
Speed
255 tok/s
First chunk
186.14s
Total response
188.11s
35

Current reasoning model

Meta - Muse Ember - Max

Meta Muse Ember Max ranks #35 in the 2026-11-11 snapshot with Intelligence Index 36, 1.05M context, $0.14 cost per task, 454 tok/s median output, and 4.94s to first chunk. It is one of 2 models at Intelligence Index 36. First chunk arrives in 4.94s, so it suits interactive loops.

Best for: Maximum-effort Meta work on hard refactors and architecture decisions, at $0.14 per task with 4.94s to first chunk.

Meta

Context
1.05M
AA Index
36
Cost per task
$0.14
Speed
454 tok/s
First chunk
4.94s
Total response
6.04s
36

Current reasoning model

SpaceXAI - Grok - 4.7 (xhigh)

SpaceXAI Grok 4.7 (xhigh) ranks #36 in the 2026-11-11 snapshot with Intelligence Index 35, 500k context, $7.77 cost per task, 63 tok/s median output, and 57.01s to first chunk. It is one of 6 models at Intelligence Index 35.

Best for: Very high-effort SpaceXAI work where a near-top index is worth a longer wait, at $7.77 per task with 57.01s to first chunk.

SpaceXAI

Context
500k
AA Index
35
Cost per task
$7.77
Speed
63 tok/s
First chunk
57.01s
Total response
64.88s
37

Current reasoning model

SpaceXAI - Grok - 4.7 (high)

SpaceXAI Grok 4.7 (high) ranks #37 in the 2026-11-11 snapshot with Intelligence Index 35, 500k context, $6.18 cost per task, 66 tok/s median output, and 36.31s to first chunk. It is one of 6 models at Intelligence Index 35.

Best for: High-effort SpaceXAI coding and multi-step agent runs, at $6.18 per task with 36.31s to first chunk.

SpaceXAI

Context
500k
AA Index
35
Cost per task
$6.18
Speed
66 tok/s
First chunk
36.31s
Total response
43.83s
38

Current reasoning model

OpenAI - GPT Sol - 6.1 (low)

OpenAI GPT Sol 6.1 (low) ranks #38 in the 2026-11-11 snapshot with Intelligence Index 35, 1M context, $0.16 cost per task, 56 tok/s median output, and 2.71s to first chunk. It is one of 6 models at Intelligence Index 35. First chunk arrives in 2.71s, so it suits interactive loops.

Best for: Light OpenAI passes and high-volume runs, at $0.16 per task with 2.71s to first chunk.

OpenAI

Context
1M
AA Index
35
Cost per task
$0.16
Speed
56 tok/s
First chunk
2.71s
Total response
11.67s
39

Current reasoning model

Anthropic - Claude Opus - 5.5 (low)

Anthropic Claude Opus 5.5 (low) ranks #39 in the 2026-11-11 snapshot with Intelligence Index 35, 1M context, $0.98 cost per task, 108 tok/s median output, and 5.10s to first chunk. It is one of 6 models at Intelligence Index 35.

Best for: Light Anthropic passes and high-volume runs, at $0.98 per task with 5.10s to first chunk.

Anthropic

Context
1M
AA Index
35
Cost per task
$0.98
Speed
108 tok/s
First chunk
5.10s
Total response
9.74s
40

Current reasoning model

Anthropic - Claude Haiku - 5.5 (xhigh)

Anthropic Claude Haiku 5.5 (xhigh) ranks #40 in the 2026-11-11 snapshot with Intelligence Index 35, 1M context, $0.25 cost per task, 236 tok/s median output, and 56.41s to first chunk. It is one of 6 models at Intelligence Index 35.

Best for: Very high-effort Anthropic work where a near-top index is worth a longer wait, at $0.25 per task with 56.41s to first chunk.

Anthropic

Context
1M
AA Index
35
Cost per task
$0.25
Speed
236 tok/s
First chunk
56.41s
Total response
58.53s
41

Current model; partial public metrics

Tencent - Hy - 4

Tencent Hy 4 ranks #41 in the 2026-11-11 snapshot with Intelligence Index 35, 1.05M context, and $1.03 cost per task. Output speed and latency are not published yet. It is one of 6 models at Intelligence Index 35.

Best for: Tencent's flagship Hy model for demanding coding work; output speed and latency are not published yet.

Tencent

Context
1.05M
AA Index
35
Cost per task
$1.03
Speed
— tok/s
First chunk
—
Total response
—
42

Current reasoning model

Meta - Muse Ember - Xhigh

Meta Muse Ember Xhigh ranks #42 in the 2026-11-11 snapshot with Intelligence Index 34, 1.05M context, $0.12 cost per task, 459 tok/s median output, and 3.83s to first chunk. It is one of 2 models at Intelligence Index 34. First chunk arrives in 3.83s, so it suits interactive loops.

Best for: Very high-effort Meta work where a near-top index is worth a longer wait, at $0.12 per task with 3.83s to first chunk.

Meta

Context
1.05M
AA Index
34
Cost per task
$0.12
Speed
459 tok/s
First chunk
3.83s
Total response
4.92s
43

Current reasoning model

Alibaba - Qwen - 3.8 Max (0902)

Alibaba Qwen 3.8 Max (0902) ranks #43 in the 2026-11-11 snapshot with Intelligence Index 34, 984k context, $10.02 cost per task, 52 tok/s median output, and 121.58s to first chunk. It is one of 2 models at Intelligence Index 34. First chunk takes 121.58s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort Alibaba work on hard refactors and architecture decisions, at $10.02 per task with 121.58s to first chunk.

Alibaba

Context
984k
AA Index
34
Cost per task
$10.02
Speed
52 tok/s
First chunk
121.58s
Total response
131.21s
44

Current reasoning model

DeepSeek - DeepSeek - V4.1 Flash 1017 (max)

DeepSeek DeepSeek V4.1 Flash 1017 (max) ranks #44 in the 2026-11-11 snapshot with Intelligence Index 33, 1M context, $0.60 cost per task, 140 tok/s median output, and 49.11s to first chunk.

Best for: Maximum-effort DeepSeek work on hard refactors and architecture decisions, at $0.60 per task with 49.11s to first chunk.

DeepSeek

Context
1M
AA Index
33
Cost per task
$0.60
Speed
140 tok/s
First chunk
49.11s
Total response
52.68s
45

Current reasoning model

Z AI - GLM - 5.3 (max)

Z AI GLM 5.3 (max) ranks #45 in the 2026-11-11 snapshot with Intelligence Index 32, 1M context, $5.01 cost per task, 69 tok/s median output, and 60.12s to first chunk. First chunk takes 60.12s, so it fits batch or async work better than interactive loops.

Best for: Maximum-effort Z AI work on hard refactors and architecture decisions, at $5.01 per task with 60.12s to first chunk.

Z AI

Context
1M
AA Index
32
Cost per task
$5.01
Speed
69 tok/s
First chunk
60.12s
Total response
67.32s
46

Current reasoning model

SpaceXAI - Grok - 4.7 (low)

SpaceXAI Grok 4.7 (low) ranks #46 in the 2026-11-11 snapshot with Intelligence Index 31, 500k context, $2.92 cost per task, 64 tok/s median output, and 13.28s to first chunk. It is one of 5 models at Intelligence Index 31.

Best for: Light SpaceXAI passes and high-volume runs, at $2.92 per task with 13.28s to first chunk.

SpaceXAI

Context
500k
AA Index
31
Cost per task
$2.92
Speed
64 tok/s
First chunk
13.28s
Total response
21.07s
47

Current reasoning model

Google - Gemini - 3.8 Flash (high)

Google Gemini 3.8 Flash (high) ranks #47 in the 2026-11-11 snapshot with Intelligence Index 31, 1M context, $7.77 cost per task, 115 tok/s median output, and 17.09s to first chunk. It is one of 5 models at Intelligence Index 31.

Best for: High-effort Google coding and multi-step agent runs, at $7.77 per task with 17.09s to first chunk.

Google

Context
1M
AA Index
31
Cost per task
$7.77
Speed
115 tok/s
First chunk
17.09s
Total response
21.45s
48

Current reasoning model

Anthropic - Claude Sonnet - 5.5 (medium)

Anthropic Claude Sonnet 5.5 (medium) ranks #48 in the 2026-11-11 snapshot with Intelligence Index 31, 1M context, $0.78 cost per task, 138 tok/s median output, and 5.07s to first chunk. It is one of 5 models at Intelligence Index 31.

Best for: Balanced Anthropic coding with a moderate task budget, at $0.78 per task with 5.07s to first chunk.

Anthropic

Context
1M
AA Index
31
Cost per task
$0.78
Speed
138 tok/s
First chunk
5.07s
Total response
8.69s
49

Current reasoning model

Thinking Machines - Inkling - 2 Small

Thinking Machines Inkling 2 Small ranks #49 in the 2026-11-11 snapshot with Intelligence Index 31, 1.02M context, $1.18 cost per task, 121 tok/s median output, and 45.32s to first chunk. It is one of 5 models at Intelligence Index 31.

Best for: Thinking Machines' flagship Inkling model for demanding coding work, at $1.18 per task with 45.32s to first chunk.

Thinking Machines

Context
1.02M
AA Index
31
Cost per task
$1.18
Speed
121 tok/s
First chunk
45.32s
Total response
49.46s
50

Current reasoning model

OpenAI - GPT Terra - 6.1 (non-reasoning)

OpenAI GPT Terra 6.1 (non-reasoning) ranks #50 in the 2026-11-11 snapshot with Intelligence Index 31, 1M context, $0.23 cost per task, 69 tok/s median output, and 2.15s to first chunk. It is one of 5 models at Intelligence Index 31. First chunk arrives in 2.15s, so it suits interactive loops.

Best for: Fast OpenAI drafts, renames, and simple edits, at $0.23 per task with 2.15s to first chunk.

OpenAI

Context
1M
AA Index
31
Cost per task
$0.23
Speed
69 tok/s
First chunk
2.15s
Total response
9.36s