Price Per TokenPrice Per Token
DeepInfravsTogether AI

DeepInfra vs Together AI

Compare pricing across 20 shared models. DeepInfra offers 60 models, Together AI offers 73.

8 Ways to Use Fewer Tokens

Get the free PDF guide — practical tips to cut your token usage and API costs. Subscribe to the Price Per Token newsletter and download instantly.

20
Shared Models
18
DeepInfra Cheaper
1
Together AI Cheaper
1
Same Price

Price Comparison — Shared Models

Model ↑DeepInfra Input Together AI Input DeepInfra Output Together AI Output Cheaper
DeepSeek V3.1 $0.210 $0.600 $0.790 $1.70DeepInfra
GLM 4.6 $0.430 $0.600 $1.74 $2.20DeepInfra
GLM 4.7 $0.400 $0.450 $1.75 $2.00DeepInfra
GLM 5 $0.800 $1.00 $2.56 $3.20DeepInfra
GPT-OSS-120b $0.150 $0.150 $0.600 $0.600Same
GPT-OSS-20b $0.030 $0.050 $0.140 $0.200DeepInfra
Kimi K2.5 $0.450 $0.500 $2.25 $2.80DeepInfra
Llama 3 8B Instruct $0.030 $0.100 $0.040 $0.100DeepInfra
Llama 3.1 8B Instruct $0.020 $0.180 $0.050 $0.180DeepInfra
Llama 3.3 70B Instruct $0.100 $0.880 $0.320 $0.880DeepInfra
Llama 4 Maverick $0.150 $0.270 $0.600 $0.850DeepInfra
meta-llama-llama-guard-4-12b $0.180 $0.200 $0.180 $0.200DeepInfra
MiniMax M2.5 $0.270 $0.300 $0.950 $1.20DeepInfra
Mistral Small 24B Instruct 2501 $0.050 $0.100 $0.080 $0.300DeepInfra
Mixtral 8x7B Instruct $0.540 $0.600 $0.540 $0.600DeepInfra
MythoMax 13B $0.400 $0.300 $0.400 $0.300Together AI
Qwen3 235B A22B Instruct 2507 $0.071 $0.200 $0.100 $0.600DeepInfra
Qwen3 235B A22B Thinking 2507 $0.230 $0.650 $2.30 $3.00DeepInfra
Qwen3 Coder 480B A35B (exacto) $0.400 $2.00 $1.60 $2.00DeepInfra
R1 0528 $0.500 $3.00 $2.15 $7.00DeepInfra

Model Coverage

Shared(20)
20
models available on both
DeepInfraTogether AI
60 total68 total

Full Provider Pricing

Frequently Asked Questions

DeepInfra is cheaper on 18 out of 20 shared models. Together AI is cheaper on 1 models. 1 models have the same price.
DeepInfra and Together AI share 20 models. DeepInfra has 40 exclusive models, while Together AI has 48 exclusive models.
Together AI offers 73 models compared to DeepInfra's 60. However, model count alone doesn't determine the better provider — consider pricing, latency, and which specific models you need.