Spreads worth shopping

Ranked by actionable price spread across distinct sellers at equal-or-better serving precision. A vendor’s own service tiers are not competition. Unpriced models are excluded. Unrated is not scored zero.

57 models

As of . Condition: equal-or-better declared precision; cheaper lower-precision is not a saving.

Worth shopping

Spread is dearest ÷ cheapest among sellers at the reference precision or better. Cheaper lower-precision endpoints do not enter this table as a saving.

ModelSpreadSellersCheapDearPrecision
Gemma 4 31B 12.38×5$0.080 OpenInference$0.990 Cerebrasbf16
gpt-oss-120b lower-quant also listed9.46×3$0.037 DeepInfra$0.350 Cerebrasbf16
GPT-5.6 Sol 5.5×3$1.00 OpenAI$5.50 Amazon Bedrockunknown
Llama 3.3 70B Instruct lower-quant also listed5.26×3$0.135 Novita$0.710 CoreWeavebf16
GPT-5.6 Sol Pro 2$1.00 OpenAI$5.00 Azureunknown
MythoMax 13B lower-quant also listed2$0.080 Parasail$0.400 DeepInfrafp16
DeepSeek V4 Pro 0423 3.33×9$0.522 StreamLake$1.74 Parasailfp8
Qwen2.5 VL 72B Instruct 3.2×2$0.250 Nebius$0.800 Parasailfp8
Qwen2.5 7B Instruct 2$0.100 Phala$0.300 Togetherunknown
GLM 5.2 2.8×15$0.500 Sail Research$1.40 Venicefp8
DeepSeek V4 Flash 0423 2.5×10$0.056 StreamLake$0.140 Mancer 2fp8
Qwen3.6 35B A3B lower-quant also listed2.5×8$0.100 DeepInfra$0.250 CoreWeavefp8
DeepSeek V3.1 lower-quant also listed2.41×5$0.270 SiliconFlow$0.650 SambaNovafp8
Qwen3 235B A22B Instruct 2507 2.29×7$0.087 GMICloud$0.200 AtlasCloudfp8
GPT-5.4 2.2×3$1.25 OpenAI$2.75 Amazon Bedrockunknown
GPT-5.5 2.2×3$2.50 OpenAI$5.50 Amazon Bedrockunknown
GPT-5.6 Luna 2.2×3$0.100 OpenAI$0.220 Amazon Bedrockunknown
GPT-5.6 Terra 2.2×3$1.00 OpenAI$2.20 Amazon Bedrockunknown
Gemini 2.5 Flash 2$0.150 Google AI Studio$0.300 Googleunknown
Gemini 2.5 Flash Lite 2$0.050 Google AI Studio$0.100 Googleunknown
Gemini 3.7 Flash 2$0.188 Google$0.375 Google AI Studiounknown
GLM 5.3 Flash 5$0.075 Z.AI$0.150 Io Netfp8
GPT-5 Mini 2$0.125 OpenAI$0.250 Azureunknown
GPT-5 Nano 2$0.025 OpenAI$0.050 Azureunknown
GPT-5.1 2$0.625 OpenAI$1.25 Azureunknown
GPT-5.2 2$0.875 OpenAI$1.75 Azureunknown
GPT-5.4 Mini 2$0.375 OpenAI$0.750 Azureunknown
GPT-5.4 Nano 2$0.100 OpenAI$0.200 Azureunknown
GPT-5.4 Pro 2$15.00 OpenAI$30.00 Azureunknown
GPT-5.6 Luna Pro 2$0.100 OpenAI$0.200 Azureunknown
GPT-5.6 Terra Pro 2$1.00 OpenAI$2.00 Azureunknown
Qwen3.6 27B 5$0.300 Chutes$0.600 CoreWeavefp8
Qwen3.5-35B-A3B 1.79×5$0.140 DeepInfra$0.250 CoreWeavefp8
DeepSeek V3.2 1.77×7$0.209 GMICloud$0.370 Alibabafp8
Llama 4 Maverick 1.75×3$0.200 DeepInfra$0.350 Parasailfp8
Qwen3 32B 1.75×3$0.080 DeepInfra$0.140 SiliconFlowfp8
Llama 3.1 Euryale 70B v2.2 1.71×2$0.850 DeepInfra$1.45 Novitafp8
GLM 5.1 1.69×13$0.910 GMICloud$1.54 Venicefp8
GLM 5 1.67×8$0.600 GMICloud$1.00 Novitafp8
Qwen3 235B A22B Thinking 2507 lower-quant also listed1.5×2$0.300 Novita$0.450 Venicefp8
R1 0528 1.4×2$0.500 SiliconFlow$0.700 Novitafp8
Qwen3.5 397B A17B lower-quant also listed1.33×4$0.450 DeepInfra$0.600 GMICloudfp8
Gemma 4 26B A4B lower-quant also listed1.3×4$0.100 NextBit$0.130 Parasailbf16
Llama 3 8B Lunaris 1.25×2$0.040 Parasail$0.050 Novitabf16
MiniMax M2.7 1.25×5$0.240 GMICloud$0.300 Minimaxfp8
MiniMax M3 lower-quant also listed1.25×8$0.240 GMICloud$0.300 Minimaxfp8
Nemotron 3.5 Lightning 1.25×2$0.080 DeepInfra$0.100 CoreWeavebf16
DeepSeek V4 Pro 0813 lower-quant also listed1.18×5$1.12 GMICloud$1.32 SiliconFlowfp8
MiniMax M2 1.18×2$0.255 Minimax$0.300 Novitafp8
GLM 4.7 Flash 1.17×2$0.060 DeepInfra$0.070 Novitabf16
GLM 4.7 lower-quant also listed1.13×3$0.480 StreamLake$0.540 Novitafp8
DeepSeek V3.1 Terminus 1.11×3$0.270 Novita$0.300 AtlasCloudfp8
Inkling Small 1.11×2$0.450 DeepInfra$0.500 BaseTenfp8
Kimi K2.7 Code lower-quant also listed1.11×3$0.859 SiliconFlow$0.950 Alibabafp8
Qwen3 Coder 480B A35B lower-quant also listed1.09×2$0.350 Venice$0.380 Novitafp8
Inkling 1.05×2$0.950 DeepInfra$1.00 BaseTenfp8
MiniMax M2.5 lower-quant also listed1.02×4$0.295 AtlasCloud$0.300 SiliconFlowfp8

Precision trades — not like-for-like savings

These look wide if you ignore quantisation. The cheap row is a lower precision; the figure is named so it is not mistaken for a shopping win.

ModelRaw ×Cheap (lower quant)Like-for-like
DeepSeek V4 Flash 073111×$0.040 Relace fp4$0.079 Morph bf16
Llama 3.1 8B Instruct11×$0.020 DeepInfra fp8$0.220 CoreWeave bf16
Qwen3 Coder 30B A3B Instruct4.18×$0.070 Novita fp8$0.070 Novita fp8
Nemotron 3 Super3.53×$0.085 DeepInfra bf16$0.085 DeepInfra bf16
MiMo-V2.53.36×$0.119 GMICloud fp8$0.400 DeepInfra bf16
MiMo-V2.5-Pro3.28×$0.304 GMICloud bf16$0.304 GMICloud bf16
Ling-3.0-flash2.86×$0.021 Novita unknown$0.060 DeepInfra bf16
Qwen3 30B A3B Instruct 25072.7×$0.048 StreamLake unknown$0.100 CoreWeave bf16
gpt-oss-20b2.5×$0.030 DeepInfra bf16$0.030 DeepInfra bf16
Llama 4 Scout2.5×$0.100 DeepInfra fp8$0.180 Novita bf16
Qwen3 Coder Next2.5×$0.120 Parasail bf16$0.120 Parasail bf16
Qwen3 VL 235B A22B Thinking2.45×$0.400 Alibaba unknown$0.980 Novita bf16
Mistral Nemo2.32×$0.019 DeepInfra fp8$0.044 Io Net fp16
Qwen3 14B2.27×$0.100 NextBit int4$0.120 DeepInfra fp8
Qwen3 VL 30B A3B Instruct2.23×$0.130 Alibaba unknown$0.200 Novita bf16
Qwen3 VL 8B Instruct2.14×$0.117 Alibaba unknown$0.250 Parasail bf16
DeepSeek V3 03242.08×$0.240 DeepInfra fp4$0.500 Crusoe bf16
Llama 3.1 70B Instruct$0.400 DeepInfra fp8$0.800 CoreWeave bf16
Nex-N2-Pro$0.250 Nex AGI fp8$0.250 Nex AGI fp8
Gemma 3 27B1.88×$0.080 DeepInfra fp8$0.119 Novita bf16
Kimi K2.61.88×$0.580 Chutes int4$0.700 Crusoe bf16
Qwen3.5-9B1.7×$0.100 DeepInfra bf16$0.100 DeepInfra bf16
Qwen3 Next 80B A3B Instruct1.67×$0.090 DeepInfra fp8$0.150 Novita bf16
Hy31.59×$0.126 GMICloud bf16$0.126 GMICloud bf16
DeepSeek V31.55×$0.257 StreamLake unknown$0.400 Novita fp8
GLM 4.5 Air1.54×$0.130 Novita bf16$0.130 Novita bf16
Qwen3.5-122B-A10B1.54×$0.260 SiliconFlow fp8$0.400 Novita bf16
Qwen3.5-27B1.54×$0.195 Alibaba unknown$0.300 Novita bf16
Qwen3 VL 235B A22B Instruct1.5×$0.200 DeepInfra fp8$0.300 Novita bf16
Qwen3 VL 30B A3B Thinking1.45×$0.200 Alibaba unknown$0.290 SiliconFlow fp8
GLM 4.61.4×$0.430 Venice fp4$0.550 Novita bf16
MiniMax M11.38×$0.400 Minimax unknown$0.550 Novita bf16
Qwen3.8 27B1.37×$0.350 Chutes fp8$0.400 AkashML bf16
Kimi K2.51.33×$0.450 DeepInfra fp4$0.540 StreamLake fp8
Kimi K31.33×$2.60 Sail Research fp4$2.85 DeepInfra bf16
Mistral Small 41.25×$0.150 Mistral unknown$0.188 Venice fp8
Nemotron 3 Ultra1.25×$0.500 DeepInfra fp4$0.625 Venice fp8
Qwen3.8 2.4T A95B1.25×$2.00 SiliconFlow fp8$2.00 SiliconFlow fp8
Mistral Small 3.2 24B1.2×$0.075 DeepInfra fp8$0.090 Parasail bf16
Muse Glimmer 30B1.17×$0.300 Parasail bf16$0.300 Parasail bf16
Trinity Large Thinking1.14×$0.220 Parasail fp4$0.220 Parasail fp4
Llama Guard 4 12B1.11×$0.180 DeepInfra bf16$0.180 DeepInfra bf16
Qwen3 30B A3B1.08×$0.120 DeepInfra fp8$0.120 DeepInfra fp8
Qwen2.5 72B Instruct1.06×$0.360 DeepInfra fp8$0.380 Novita bf16
Llama 3.2 3B Instruct1.02×$0.050 Parasail bf16$0.050 Parasail bf16

Excluded from the ranked list

  • 0 unpriced — no finite input price (missing price is not free).
  • 171 single-seller — one shop is not a market.
  • 47 no actionable equal-or-better spread.
  • 45 precision trades — listed above, not as savings.
  • 0 publication denylist (FakeProvider / Stealth) — never shown.