← All Tools

Best Llama 4 Scout Alternatives

Llama 4 Scout by Meta is a open source model priced at $0.18/0.63 per 1M tokens (in/out). It's already affordable, but you might want different strengths or features.

Llama 4 Scout

MetaOpen Source

Input

$0.18/1M

Output

$0.63/1M

Context

10M

Max Output

32K

Why Switch from Llama 4 Scout?

Lower benchmarks than Maverick
Limited tool-use support

Top Alternatives

#1Llama 4 MaverickMetaOpen Source

Higher benchmark scores, open-source and self-hostable.

Input

$0.31/1M

72% more

Output

$0.85/1M

35% more

Context

1M

Max Output

32K

MMLU-Pro: 80.5%(+6.3%)HumanEval: 90.2%(+4.2%)
#2DeepSeek V3DeepSeekOpen Source

48% cheaper, higher benchmark scores, adds reasoning.

Input

$0.14/1M

22% cheaper

Output

$0.28/1M

56% cheaper

Context

164K

Max Output

16K

MMLU-Pro: 78%(+3.8%)HumanEval: 89%(+3.0%)
#3MiniMax M2.5MiniMaxOpen Source

Higher benchmark scores, 128K max output, adds reasoning.

Input

$0.3/1M

67% more

Output

$1.2/1M

90% more

Context

200K

Max Output

128K

MMLU-Pro: 82%(+7.8%)HumanEval: 90%(+4.0%)
#4Gemini 2.5 FlashGoogleBudget

7% cheaper, higher benchmark scores, 66K max output.

Input

$0.15/1M

17% cheaper

Output

$0.6/1M

5% cheaper

Context

1M

Max Output

66K

MMLU-Pro: 76%(+1.8%)HumanEval: 89.5%(+3.5%)
#5GPT-4o MiniOpenAIBudget

7% cheaper, adds tool-use.

Input

$0.15/1M

17% cheaper

Output

$0.6/1M

5% cheaper

Context

128K

Max Output

16K

MMLU-Pro: 68%(-6.2%)HumanEval: 87.2%(+1.2%)
#6Mistral Medium 3MistralMid-Tier

Comparable performance, adds tool-use.

Input

$0.4/1M

122% more

Output

$2/1M

217% more

Context

128K

Max Output

16K

MMLU-Pro: 76%(+1.8%)HumanEval: 87%(+1.0%)
#7o3OpenAIReasoning

Higher benchmark scores, 100K max output, adds tool-use, reasoning.

Input

$0.4/1M

122% more

Output

$1.6/1M

154% more

Context

200K

Max Output

100K

MMLU-Pro: 87%(+12.8%)HumanEval: 94.5%(+8.5%)
#8GLM-4.7Zhipu AIMid-Tier

Higher benchmark scores, 128K max output, adds reasoning.

Input

$0.6/1M

233% more

Output

$2.2/1M

249% more

Context

200K

Max Output

128K

MMLU-Pro: 84.3%(+10.1%)HumanEval:

Full Comparison Table

ModelInput $/1MOutput $/1MContextMMLU-ProHumanEvalScore
Llama 4 MaverickMeta$0.3172% more$0.8535% more1M80.5%+6.3%90.2%+4.2%90
DeepSeek V3DeepSeek$0.1422% cheaper$0.2856% cheaper164K78%+3.8%89%+3.0%85
MiniMax M2.5MiniMax$0.3067% more$1.2090% more200K82%+7.8%90%+4.0%75
Gemini 2.5 FlashGoogle$0.1517% cheaper$0.605% cheaper1M76%+1.8%89.5%+3.5%73
GPT-4o MiniOpenAI$0.1517% cheaper$0.605% cheaper128K68%-6.2%87.2%+1.2%71
Mistral Medium 3Mistral$0.40122% more$2.00217% more128K76%+1.8%87%+1.0%68
o3OpenAI$0.40122% more$1.60154% more200K87%+12.8%94.5%+8.5%63
GLM-4.7Zhipu AI$0.60233% more$2.20249% more200K84.3%+10.1%58
o4-miniOpenAI$1.10511% more$4.40598% more200K85%+10.8%93.5%+7.5%53
Gemini 3 FlashGoogle$0.50178% more$3.00376% more1M78%+3.8%90%+4.0%53
Mistral Large 3Mistral$2.001011% more$5.00694% more128K83%+8.8%91%+5.0%53
Claude Haiku 4.5Anthropic$0.80344% more$4.00535% more200K69.4%-4.8%88.1%+2.1%51
DeepSeek R1DeepSeek$0.55206% more$2.19248% more128K84%+9.8%92%+6.0%50
GLM-5Zhipu AI$1.00456% more$3.20408% more200K70.4%-3.8%91%+5.0%46
Claude Opus 4.6Anthropic$5.002678% more$25.003868% more200K89.5%+15.3%95%+9.0%43
Claude Sonnet 4.6Anthropic$3.001567% more$15.002281% more200K86%+11.8%94%+8.0%43
Claude Sonnet 4.5Anthropic$3.001567% more$15.002281% more200K84.5%+10.3%93%+7.0%43
GPT-5.3 CodexOpenAI$2.001011% more$16.002440% more200K90%+15.8%96.5%+10.5%43
GPT-5.2 CodexOpenAI$1.75872% more$14.002122% more200K89%+14.8%95.5%+9.5%43
GPT-5OpenAI$1.25594% more$10.001487% more128K88.5%+14.3%95%+9.0%40
Gemini 3.1 ProGoogle$2.001011% more$12.001805% more1M91%+16.8%95%+9.0%40
Gemini 3 ProGoogle$2.001011% more$12.001805% more1M89.8%+15.6%94%+8.0%40
Gemini 2.5 ProGoogle$1.25594% more$10.001487% more1M87.5%+13.3%93.5%+7.5%40
Grok 4xAI$3.001567% more$15.002281% more128K86%+11.8%93%+7.0%40
GPT-4oOpenAI$2.501289% more$10.001487% more128K80.5%+6.3%91%+5.0%36

Head-to-Head Comparisons

Alternatives for Other Models