← All Tools

Best GLM-5 Alternatives

GLM-5 by Zhipu AI is a flagship model priced at $1/3.2 per 1M tokens (in/out). Looking for a better deal or different capabilities? Here are the best options.

GLM-5

Zhipu AIFlagship

Input

$1/1M

Output

$3.2/1M

Context

200K

Max Output

128K

Why Switch from GLM-5?

MMLU-Pro lags behind Western flagships
744B parameters — heavy to self-host
China-based — availability concerns

Top Alternatives

#1Mistral Large 3MistralFlagship

Higher benchmark scores.

Input

$2/1M

100% more

Output

$5/1M

56% more

Context

128K

Max Output

16K

MMLU-Pro: 83%(+12.6%)HumanEval: 91%(Same)GPQA:
#2GPT-5OpenAIFlagship

Higher benchmark scores, adds audio.

Input

$1.25/1M

25% more

Output

$10/1M

213% more

Context

128K

Max Output

16K

MMLU-Pro: 88.5%(+18.1%)HumanEval: 95%(+4.0%)GPQA: 73.5%(+1.5%)
#3o3OpenAIReasoning

Dramatically cheaper (52% less), higher benchmark scores.

Input

$0.4/1M

60% cheaper

Output

$1.6/1M

50% cheaper

Context

200K

Max Output

100K

MMLU-Pro: 87%(+16.6%)HumanEval: 94.5%(+3.5%)GPQA: 79.2%(+7.2%)
#4Gemini 3 FlashGoogleBudget

17% cheaper, higher benchmark scores, 1M context (5x more).

Input

$0.5/1M

50% cheaper

Output

$3/1M

6% cheaper

Context

1M

Max Output

66K

MMLU-Pro: 78%(+7.6%)HumanEval: 90%(-1.0%)GPQA:
#5Claude Opus 4.6AnthropicFlagship

Higher benchmark scores.

Input

$5/1M

400% more

Output

$25/1M

681% more

Context

200K

Max Output

32K

MMLU-Pro: 89.5%(+19.1%)HumanEval: 95%(+4.0%)GPQA: 75.5%(+3.5%)
#6GPT-5.3 CodexOpenAIFlagship

Higher benchmark scores.

Input

$2/1M

100% more

Output

$16/1M

400% more

Context

200K

Max Output

66K

MMLU-Pro: 90%(+19.6%)HumanEval: 96.5%(+5.5%)GPQA: 78%(+6.0%)
#7GPT-5.2 CodexOpenAIFlagship

Higher benchmark scores.

Input

$1.75/1M

75% more

Output

$14/1M

338% more

Context

200K

Max Output

66K

MMLU-Pro: 89%(+18.6%)HumanEval: 95.5%(+4.5%)GPQA: 76%(+4.0%)
#8GLM-4.7Zhipu AIMid-Tier

33% cheaper, higher benchmark scores.

Input

$0.6/1M

40% cheaper

Output

$2.2/1M

31% cheaper

Context

200K

Max Output

128K

MMLU-Pro: 84.3%(+13.9%)HumanEval: GPQA: 85.7%(+13.7%)

Full Comparison Table

ModelInput $/1MOutput $/1MContextMMLU-ProHumanEvalScore
Mistral Large 3Mistral$2.00100% more$5.0056% more128K83%+12.6%91%Same90
GPT-5OpenAI$1.2525% more$10.00213% more128K88.5%+18.1%95%+4.0%85
o3OpenAI$0.4060% cheaper$1.6050% cheaper200K87%+16.6%94.5%+3.5%85
Gemini 3 FlashGoogle$0.5050% cheaper$3.006% cheaper1M78%+7.6%90%-1.0%85
Claude Opus 4.6Anthropic$5.00400% more$25.00681% more200K89.5%+19.1%95%+4.0%80
GPT-5.3 CodexOpenAI$2.00100% more$16.00400% more200K90%+19.6%96.5%+5.5%80
GPT-5.2 CodexOpenAI$1.7575% more$14.00338% more200K89%+18.6%95.5%+4.5%80
GLM-4.7Zhipu AI$0.6040% cheaper$2.2031% cheaper200K84.3%+13.9%79
Mistral Medium 3Mistral$0.4060% cheaper$2.0038% cheaper128K76%+5.6%87%-4.0%79
o4-miniOpenAI$1.1010% more$4.4038% more200K85%+14.6%93.5%+2.5%75
Gemini 3.1 ProGoogle$2.00100% more$12.00275% more1M91%+20.6%95%+4.0%75
Gemini 3 ProGoogle$2.00100% more$12.00275% more1M89.8%+19.4%94%+3.0%75
Gemini 2.5 FlashGoogle$0.1585% cheaper$0.6081% cheaper1M76%+5.6%89.5%-1.5%75
Grok 4xAI$3.00200% more$15.00369% more128K86%+15.6%93%+2.0%75
DeepSeek R1DeepSeek$0.5545% cheaper$2.1932% cheaper128K84%+13.6%92%+1.0%73
MiniMax M2.5MiniMax$0.3070% cheaper$1.2063% cheaper200K82%+11.6%90%-1.0%73
Gemini 2.5 ProGoogle$1.2525% more$10.00213% more1M87.5%+17.1%93.5%+2.5%70
Claude Haiku 4.5Anthropic$0.8020% cheaper$4.0025% more200K69.4%-1.0%88.1%-2.9%69
Claude Sonnet 4.6Anthropic$3.00200% more$15.00369% more200K86%+15.6%94%+3.0%65
Claude Sonnet 4.5Anthropic$3.00200% more$15.00369% more200K84.5%+14.1%93%+2.0%65
GPT-4oOpenAI$2.50150% more$10.00213% more128K80.5%+10.1%91%Same65
Llama 4 MaverickMeta$0.3169% cheaper$0.8573% cheaper1M80.5%+10.1%90.2%-0.8%63
Llama 4 ScoutMeta$0.1882% cheaper$0.6380% cheaper10M74.2%+3.8%86%-5.0%63
DeepSeek V3DeepSeek$0.1486% cheaper$0.2891% cheaper164K78%+7.6%89%-2.0%63
GPT-4o MiniOpenAI$0.1585% cheaper$0.6081% cheaper128K68%-2.4%87.2%-3.8%62

Head-to-Head Comparisons

Alternatives for Other Models