Updated Oct 16, 10:22 PM GMT+2

Model Analytics

Usage of models across Spettro. See which models are winning real usage, how the mix shifts over time, and where momentum is moving each week.

SuperFast (DeepSeek V4 Flash) leads with 42% of token usage, ahead of Kimi K3 at 18% and SuperSmart (DeepSeek V4 Pro) at 13.1%.

Total Tokens 56.8B Jun 2026 to Oct 2026

Top Model 24h Velo 3.1 Flash 26.2M tokens · 76.2%

Top models.

Usage of models across Spettro.

SuperFast (DeepSeek V4 Flash) DeepSeek

23.9B

-71%

Kimi K3 Moonshot

10.2B

-31%

SuperSmart (DeepSeek V4 Pro) DeepSeek

7.4B

-69%

GLM 5.2 Zhipu

+721%

MiniMax M3 MiniMax

4.5B

-74%

Kimi K2.7 Coder Moonshot

3.9B

+13%

Qwen3.8 Max Qwen

1.3B

+4%

Velo 3.1 Flash Stealth

329.6M

New

Qwen3.7 Max Qwen

118.9M

New

GLM 5.3 Zhipu

78.8M

-83%

Qwen3.6 Flash Qwen

39M

New

Qwen3.7 Plus Qwen

5.9M

New

GLM 5.1 Zhipu

67.3K

New

Leaderboard.

Full ranking by token usage.

#		Model	Provider	Tokens	Market Share	Δ 7d
01	=	SuperFast (DeepSeek V4 Flash)	DeepSeek	23.9B	42%	-71%
02	▼2	Kimi K3	Moonshot	10.2B	18%	-31%
03	=	SuperSmart (DeepSeek V4 Pro)	DeepSeek	7.4B	13.1%	-69%
04	·	GLM 5.2	Zhipu	5B	8.9%	+721%
05	▼1	MiniMax M3	MiniMax	4.5B	7.9%	-74%
06	▼1	Kimi K2.7 Coder	Moonshot	3.9B	6.8%	+13%
07	▼1	Qwen3.8 Max	Qwen	1.3B	2.4%	+4%
08	·	Velo 3.1 Flash	Stealth	329.6M	0.6%	New
09	·	Qwen3.7 Max	Qwen	118.9M	0.2%	New
10	·	GLM 5.3	Zhipu	78.8M	0.1%	-83%
11	·	Qwen3.6 Flash	Qwen	39M	0.1%	New
12	·	Qwen3.7 Plus	Qwen	5.9M	0%	New
13	·	GLM 5.1	Zhipu	67.3K	0%	New

Session cost.

Average cost per session.

COST / SESSIONTOKENS / SESSION

$0.002833SuperFast (DeepSeek V4 Flash)

$0.004401Qwen3.6 Flash

$0.008843GLM 5.1

$0.0141Kimi K2.7 Coder

$0.0141Qwen3.8 Max

$0.0151MiniMax M3

$0.0181GLM 5.3

$0.0197SuperSmart (DeepSeek V4 Pro)

$0.0254Qwen3.7 Plus

$0.0348GLM 5.2

$0.0383Kimi K3

$0.0463Qwen3.7 Max

Token cost.

Fixed pricing per 1M tokens (input / output).

COST / 1M TOKENS (IN → OUT)

$0.450 / $1.350SuperFast (DeepSeek V4 Flash)

$0.600 / $2.400Minimax M3

$0.800 / $3.200Qwen3.7 Plus

$0.894 / $3.713Kimi K2.7 Code

$0.660 / $3.961Qwen3.6 Flash

$1.320 / $3.840SuperSmart (DeepSeek V4 Pro)

$1.230 / $4.150GLM 5.3

$1.720 / $5.050Qwen3.8 Max

$3.500 / $14.500Kimi K3

Cache hits.

Share of prompt tokens served from cache.

94.9% of prompt tokens were
served from cache

94.5% median model hit rate

9 models reporting

Cached share of prompt tokens

96.7%GLM 5.2

95.5%SuperFast (DeepSeek V4 Flash)

95.1%Qwen3.8 Max

94.9%SuperSmart (DeepSeek V4 Pro)

94.5%Kimi K3

94.1%MiniMax M3

93.7%GLM 5.3

92.7%Kimi K2.7 Coder

92.3%Velo 3.1 Flash

[*] dashed marker = fleet average · only requests with cached tokens are counted; models without cache hits are hidden.