← All Tools

Best Gemini 3 Flash Alternatives

Gemini 3 Flash by Google is a budget model priced at $0.5/3 per 1M tokens (in/out). Looking for a better deal or different capabilities? Here are the best options.

Gemini 3 Flash

GoogleBudget

Input

$0.5/1M

Output

$3/1M

Context

1M

Max Output

66K

Why Switch from Gemini 3 Flash?

Less capable than Pro on complex reasoning
Audio input priced separately ($1/1M)

Top Alternatives

#1o3OpenAIReasoning

43% cheaper, higher benchmark scores.

Input

$0.4/1M

20% cheaper

Output

$1.6/1M

47% cheaper

Context

200K

Max Output

100K

MMLU-Pro: 87%(+9.0%)HumanEval: 94.5%(+4.5%)
#2Gemini 2.5 FlashGoogleBudget

Dramatically cheaper (79% less), comparable performance.

Input

$0.15/1M

70% cheaper

Output

$0.6/1M

80% cheaper

Context

1M

Max Output

66K

MMLU-Pro: 76%(-2.0%)HumanEval: 89.5%(-0.5%)
#3GLM-4.7Zhipu AIMid-Tier

20% cheaper, comparable performance.

Input

$0.6/1M

20% more

Output

$2.2/1M

27% cheaper

Context

200K

Max Output

128K

MMLU-Pro: 84.3%(+6.3%)HumanEval:
#4o4-miniOpenAIReasoning

Comparable performance.

Input

$1.1/1M

120% more

Output

$4.4/1M

47% more

Context

200K

Max Output

100K

MMLU-Pro: 85%(+7.0%)HumanEval: 93.5%(+3.5%)
#5Mistral Large 3MistralFlagship

Higher benchmark scores.

Input

$2/1M

300% more

Output

$5/1M

67% more

Context

128K

Max Output

16K

MMLU-Pro: 83%(+5.0%)HumanEval: 91%(+1.0%)
#6MiniMax M2.5MiniMaxOpen Source

Dramatically cheaper (57% less), comparable performance, open-source and self-hostable.

Input

$0.3/1M

40% cheaper

Output

$1.2/1M

60% cheaper

Context

200K

Max Output

128K

MMLU-Pro: 82%(+4.0%)HumanEval: 90%(Same)
#7Mistral Medium 3MistralMid-Tier

31% cheaper.

Input

$0.4/1M

20% cheaper

Output

$2/1M

33% cheaper

Context

128K

Max Output

16K

MMLU-Pro: 76%(-2.0%)HumanEval: 87%(-3.0%)
#8Claude Haiku 4.5AnthropicBudget

Same category, different trade-offs.

Input

$0.8/1M

60% more

Output

$4/1M

33% more

Context

200K

Max Output

8K

MMLU-Pro: 69.4%(-8.6%)HumanEval: 88.1%(-1.9%)

Full Comparison Table

ModelInput $/1MOutput $/1MContextMMLU-ProHumanEvalScore
o3OpenAI$0.4020% cheaper$1.6047% cheaper200K87%+9.0%94.5%+4.5%85
Gemini 2.5 FlashGoogle$0.1570% cheaper$0.6080% cheaper1M76%-2.0%89.5%-0.5%83
GLM-4.7Zhipu AI$0.6020% more$2.2027% cheaper200K84.3%+6.3%79
o4-miniOpenAI$1.10120% more$4.4047% more200K85%+7.0%93.5%+3.5%75
Mistral Large 3Mistral$2.00300% more$5.0067% more128K83%+5.0%91%+1.0%75
MiniMax M2.5MiniMax$0.3040% cheaper$1.2060% cheaper200K82%+4.0%90%Same73
Mistral Medium 3Mistral$0.4020% cheaper$2.0033% cheaper128K76%-2.0%87%-3.0%72
Claude Haiku 4.5Anthropic$0.8060% more$4.0033% more200K69.4%-8.6%88.1%-1.9%69
GPT-4o MiniOpenAI$0.1570% cheaper$0.6080% cheaper128K68%-10.0%87.2%-2.8%69
DeepSeek R1DeepSeek$0.5510% more$2.1927% cheaper128K84%+6.0%92%+2.0%66
Claude Opus 4.6Anthropic$5.00900% more$25.00733% more200K89.5%+11.5%95%+5.0%65
GPT-5.3 CodexOpenAI$2.00300% more$16.00433% more200K90%+12.0%96.5%+6.5%65
GPT-5.2 CodexOpenAI$1.75250% more$14.00367% more200K89%+11.0%95.5%+5.5%65
Llama 4 MaverickMeta$0.3138% cheaper$0.8572% cheaper1M80.5%+2.5%90.2%+0.2%63
GPT-5OpenAI$1.25150% more$10.00233% more128K88.5%+10.5%95%+5.0%60
Gemini 3.1 ProGoogle$2.00300% more$12.00300% more1M91%+13.0%95%+5.0%60
Gemini 3 ProGoogle$2.00300% more$12.00300% more1M89.8%+11.8%94%+4.0%60
Gemini 2.5 ProGoogle$1.25150% more$10.00233% more1M87.5%+9.5%93.5%+3.5%60
GLM-5Zhipu AI$1.00100% more$3.207% more200K70.4%-7.6%91%+1.0%60
Claude Sonnet 4.6Anthropic$3.00500% more$15.00400% more200K86%+8.0%94%+4.0%58
Claude Sonnet 4.5Anthropic$3.00500% more$15.00400% more200K84.5%+6.5%93%+3.0%58
Llama 4 ScoutMeta$0.1864% cheaper$0.6379% cheaper10M74.2%-3.8%86%-4.0%56
DeepSeek V3DeepSeek$0.1472% cheaper$0.2891% cheaper164K78%Same89%-1.0%56
Grok 4xAI$3.00500% more$15.00400% more128K86%+8.0%93%+3.0%53
GPT-4oOpenAI$2.50400% more$10.00233% more128K80.5%+2.5%91%+1.0%40

Head-to-Head Comparisons

Alternatives for Other Models