Compare up to 4 models side-by-side — pricing, benchmarks, and capabilities. View all models →
Models Compare
Model Comparison Tool

Compare Models
Side-by-Side

Select up to 4 models. Compare pricing, context, benchmarks, capabilities, and deployment options. Use the cost estimator to find the best value for your use case.

💰 Pricing
📏 Context
🖼 Modalities
⚙️ Capabilities
📊 Benchmarks
🚀 Deployment
Comparing 3 models
Highlight winner:
Last updated: March 9, 2026
⚡ Serverless Input
Per 1M tokens
$3.00 👑
$1.25
$0.27
⚡ Serverless Output
Per 1M tokens
$15.00
$10.00
$1.10 👑
Standard Input
Direct API / per 1M
$3.00 👑
$1.25
$0.27
Standard Output
Direct API / per 1M
$15.00
$10.00
$1.10 👑
Free Tier
limited
Prompt Caching
Write / Read per 1M
$3.75 / $0.30
$3.13 / $0.31
Batch Discount
−50%
−50%
Providers
Where available
11 providers
8 providers
6 providers
Context Window
200K
1M 👑
64K
Max Output Tokens
16,384 👑
8,192
8,192
Throughput (avg)
Tokens / second
85 t/s
110 t/s 👑
120 t/s
TTFT P50
Time to first token
430ms
510ms
280ms 👑
Uptime (30d)
99.98%
99.95%
99.90%
Training Cutoff
Jan 2026 👑
Jan 2026 👑
Oct 2024
Parameter Count
~200B est.
Undisclosed
671B (MoE) MoE
Text Input
Image Input
Vision
video too
Audio Input
File / PDF Input
Text Output
Code Output
Embedding Output
Tools / Function Calling
JSON Mode
Structured Outputs
Streaming
System Prompt
Reasoning Tokens
Extended thinking
Partial
✓ 👑
Partial
FIM Completion
Code infill
Distillable
Fine-Tuning
Zero Data Retention
ZDR
MMLU
General knowledge
88.7
89.2 👑
86.7
HumanEval
Code generation
92.4 👑
88.1
87.6
GSM8K
Math reasoning
96.2 👑
95.8
93.5
GPQA Diamond
Expert reasoning
68.3
72.4 👑
62.1
MT-Bench
Multi-turn quality
9.1 👑
8.8
8.7
MATH
Competition math
81.5 👑
80.0
79.2
Benchmark Wins
4 / 6 👑
2 / 6
0 / 6
Open Weights
Serverless Deploy
Private Deploy
Your infra
via key routing
self-host
Air-Gap Ready
Metaprise
Via substitute
Via substitute
native
Zero Data Retention
HIPAA Ready
SOC 2 Type II
GDPR Compliant
AuditChain Logging
★ Metaprise
License
Commercial API
Commercial API
MIT (Open)
💡 Cost Estimator
Enter your expected usage to compare real costs across all selected models.
Metric
🤖 Claude Sonnet 4.6
Gemini 2.5 Pro
DeepSeek V3.2
Daily Cost
$36.00
$14.00 −61%
$5.74 −84%
Monthly (30d)
$1,080
$420
$172
Annual (365d)
$13,140
$5,110
$2,095
Per Request (avg)
$0.0072
$0.0028
$0.00115
vs. Most Expensive
baseline
Save 61%
Save 84%
⚠️ Estimates based on listed Serverless pricing. Actual costs may vary with caching, batch discounts, and provider selection. Contact enterprise sales for custom pricing →