Live data — the endpoint grid

The endpoint grid

219 endpoints, one row per model, provider, and pin, as seen on the most recent sweep. Prices in dollars per million tokens.

model	provider	pin	quant	$/M in	$/M out	err rate	last seen
deepseek-v3.2	Alibaba	alibaba/fp8	fp8	0.370	1.111	0.0%	08-31 21:45Z
deepseek-v3.2	AtlasCloud	atlas-cloud/fp8	fp8	0.260	0.380	4.7%	08-31 21:45Z
deepseek-v3.2	Baidu	baidu/fp8	fp8	0.280	0.420	14.0%	08-31 21:45Z
deepseek-v3.2	DeepInfra	deepinfra/fp4	fp4	0.260	0.380	20.9%	08-31 21:45Z
deepseek-v3.2	DigitalOcean	digitalocean	unknown	0.250	0.800	14.0%	08-31 21:45Z
deepseek-v3.2	Friendli	friendli	unknown	0.500	1.500	39.5%	08-31 21:45Z
deepseek-v3.2	GMICloud	gmicloud/fp8	fp8	0.209	0.310	0.0%	08-31 21:45Z
deepseek-v3.2	Google	google-vertex	unknown	0.560	1.680	37.2%	08-31 21:45Z
deepseek-v3.2	Mara	mara	unknown	3.000	4.500	14.0%	08-31 21:45Z
deepseek-v3.2	Novita	novita/fp8	fp8	0.269	0.400	0.0%	08-31 21:45Z
deepseek-v3.2	Phala	phala	unknown	1.000	1.000	0.0%	08-31 21:45Z
deepseek-v3.2	SambaNova	sambanova	unknown	3.000	4.500	14.0%	08-31 21:45Z
deepseek-v3.2	SiliconFlow	siliconflow/fp8	fp8	0.259	0.420	0.0%	08-31 21:45Z
deepseek-v3.2	StreamLake	streamlake/fp8	fp8	0.214	0.322	4.7%	08-31 21:45Z
deepseek-v3.2	Venice	venice	unknown	0.330	0.480	34.9%	08-31 21:45Z
deepseek-v4-flash	Alibaba	alibaba/fp8	fp8	0.134	0.268	0.0%	08-31 21:45Z
deepseek-v4-flash	AtlasCloud	atlas-cloud/fp4	fp4	0.140	0.280	0.0%	08-31 21:45Z
deepseek-v4-flash	Azure	azure/us	unknown	0.210	0.560	0.0%	08-31 21:45Z
deepseek-v4-flash	Baidu	baidu/fp8	fp8	0.081	0.162	67.4%	08-31 21:45Z
deepseek-v4-flash	Cloudflare	cloudflare	unknown	0.440	1.320	0.0%	08-31 21:45Z
deepseek-v4-flash	CoreWeave	coreweave/fp8	fp8	0.140	0.280	0.0%	08-31 21:45Z
deepseek-v4-flash	DeepInfra	deepinfra/fp8	fp8	0.090	0.180	95.3%	08-31 21:45Z
deepseek-v4-flash	DigitalOcean	digitalocean	unknown	0.068	0.168	0.0%	08-31 21:45Z
deepseek-v4-flash	GMICloud	gmicloud/fp8	fp8	0.112	0.224	0.0%	08-31 21:45Z
deepseek-v4-flash	Mancer 2	mancer/fp8	fp8	0.180	0.500	2.3%	08-31 21:45Z
deepseek-v4-flash	NextBit	nextbit/fp8	fp8	0.140	0.280	0.0%	08-31 21:45Z
deepseek-v4-flash	Novita	novita/fp8	fp8	0.140	0.280	0.0%	08-31 21:45Z
deepseek-v4-flash	Parasail	parasail/fp8	fp8	0.140	0.280	0.0%	08-31 21:45Z
deepseek-v4-flash	Phala	phala	unknown	0.200	0.400	0.0%	08-31 21:45Z
deepseek-v4-flash	SiliconFlow	siliconflow/fp8	fp8	0.130	0.280	0.0%	08-31 21:45Z
deepseek-v4-flash	StreamLake	streamlake/fp8	fp8	0.081	0.162	0.0%	08-31 21:45Z
deepseek-v4-flash	Venice	venice	unknown	0.138	0.275	0.0%	08-31 21:45Z
deepseek-v4-pro	Alibaba	alibaba/fp8	fp8	1.416	2.832	0.0%	08-31 21:45Z
deepseek-v4-pro	AtlasCloud	atlas-cloud/fp4	fp4	1.680	3.380	0.0%	08-31 21:45Z
deepseek-v4-pro	Azure	azure/us	unknown	1.910	3.830	0.0%	08-31 21:45Z
deepseek-v4-pro	Baidu	baidu/fp8	fp8	0.869	1.737	0.0%	08-31 21:45Z
deepseek-v4-pro	BaseTen	baseten/fp4	fp4	1.740	3.480	0.0%	08-31 21:45Z
deepseek-v4-pro	CoreWeave	coreweave/fp8	fp8	1.150	2.550	62.8%	08-31 21:45Z
deepseek-v4-pro	DeepInfra	deepinfra/fp8	fp8	1.300	2.600	9.3%	08-31 21:45Z
deepseek-v4-pro	DigitalOcean	digitalocean	unknown	0.870	1.740	4.7%	08-31 21:45Z
deepseek-v4-pro	Fireworks	fireworks	unknown	1.740	3.480	100.0%	08-31 21:45Z
deepseek-v4-pro	GMICloud	gmicloud/fp8	fp8	1.044	2.088	9.3%	08-31 21:45Z
deepseek-v4-pro	Ionstream	ionstream/fp4	fp4	1.131	2.262	0.0%	08-31 21:45Z
deepseek-v4-pro	NextBit	nextbit/fp8	fp8	1.720	3.450	0.0%	08-31 21:45Z
deepseek-v4-pro	Novita	novita/fp8	fp8	1.600	3.200	0.0%	08-31 21:45Z
deepseek-v4-pro	Parasail	parasail/fp8	fp8	1.740	3.480	0.0%	08-31 21:45Z
deepseek-v4-pro	SiliconFlow	siliconflow/fp8	fp8	1.502	3.135	0.0%	08-31 21:45Z
deepseek-v4-pro	StreamLake	streamlake/fp8	fp8	1.028	2.055	0.0%	08-31 21:45Z
deepseek-v4-pro	Venice	venice	unknown	1.650	3.301	0.0%	08-31 21:45Z
gemma4-31b	Cerebras	cerebras/fp16	fp16	0.990	1.490	0.0%	08-31 21:45Z
gemma4-31b	Chutes	chutes/fp4	fp4	0.120	0.370	0.0%	08-31 21:45Z
gemma4-31b	CoreWeave	coreweave/fp4	fp4	0.100	0.340	0.0%	08-31 21:45Z
gemma4-31b	Crusoe	crusoe	unknown	0.140	0.400	0.0%	08-31 21:45Z
gemma4-31b	DeepInfra	deepinfra/fp8	fp8	0.130	0.380	0.0%	08-31 21:45Z
gemma4-31b	DeepInfra	deepinfra/turbo	fp4	0.090	0.340	14.0%	08-31 21:45Z
gemma4-31b	DeepInfra	deepinfra/ultra	fp8	0.270	0.760	0.0%	08-31 21:45Z
gemma4-31b	Friendli	friendli	unknown	0.140	0.400	0.0%	08-31 21:45Z
gemma4-31b	ModelRun	modelrun/fp4	fp4	0.750	1.000	0.0%	08-31 21:45Z
gemma4-31b	Novita	novita/bf16	bf16	0.140	0.400	0.0%	08-31 21:45Z
gemma4-31b	Parasail	parasail/fp8	fp8	0.150	0.400	0.0%	08-31 21:45Z
gemma4-31b	Phala	phala	unknown	0.150	0.460	0.0%	08-31 21:45Z
gemma4-31b	SambaNova	sambanova	unknown	0.380	1.150	0.0%	08-31 21:45Z
gemma4-31b	SiliconFlow	siliconflow/fp8	fp8	0.130	0.400	0.0%	08-31 21:45Z
gemma4-31b	Together	together	unknown	0.390	0.970	0.0%	08-31 21:45Z
gemma4-31b	Venice	venice/bf16	bf16	0.120	0.360	0.0%	08-31 21:45Z
glm-5.2	Alibaba	alibaba/fast	fp8	2.310	7.260	0.0%	08-31 21:45Z
glm-5.2	Alibaba	alibaba/fp8	fp8	0.966	3.036	0.0%	08-31 21:45Z
glm-5.2	Ambient	ambient/fp8	fp8	0.600	2.000	0.0%	08-31 21:45Z
glm-5.2	AtlasCloud	atlas-cloud/fp8	fp8	1.260	3.960	0.0%	08-31 21:45Z
glm-5.2	Baidu	baidu/fp8	fp8	0.419	1.316	0.0%	08-31 21:45Z
glm-5.2	BaseTen	baseten/fast	fp8	2.100	6.600	0.0%	08-31 21:45Z
glm-5.2	BaseTen	baseten/fp8	fp8	1.400	4.400	2.3%	08-31 21:45Z
glm-5.2	Cloudflare	cloudflare	unknown	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	CoreWeave	coreweave/fp4	fp4	0.760	2.420	23.3%	08-31 21:45Z
glm-5.2	Crusoe	crusoe/fp8	fp8	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	Decart	decart/fp4	fp4	0.684	2.280	0.0%	08-31 21:45Z
glm-5.2	DeepInfra	deepinfra/fp4	fp4	0.487	1.560	0.0%	08-31 21:45Z
glm-5.2	DigitalOcean	digitalocean	unknown	0.700	2.200	0.0%	08-31 21:45Z
glm-5.2	Fireworks	fireworks	unknown	1.400	4.400	4.7%	08-31 21:45Z
glm-5.2	Fireworks	fireworks/fast	unknown	2.100	6.600	7.0%	08-31 21:45Z
glm-5.2	Fireworks	fireworks/fast-us	unknown	2.100	6.600	9.3%	08-31 21:45Z
glm-5.2	Friendli	friendli	unknown	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	GMICloud	gmicloud/fp8	fp8	1.050	3.300	0.0%	08-31 21:45Z
glm-5.2	Inceptron	inceptron/fp4	fp4	0.710	2.350	0.0%	08-31 21:45Z
glm-5.2	Mistral	mistral	unknown	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	Mistral	mistral/eu	unknown	1.540	4.840	0.0%	08-31 21:45Z
glm-5.2	Mistral	mistral/zdr	unknown	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	Novita	novita/fp8	fp8	0.742	2.332	0.0%	08-31 21:45Z
glm-5.2	Parasail	parasail/fp4	fp4	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	Phala	phala/fp8	fp8	1.260	3.000	0.0%	08-31 21:45Z
glm-5.2	Reka	reka/fp8	fp8	1.100	3.750	0.0%	08-31 21:45Z
glm-5.2	Sail Research	sail-research/fp8	fp8	0.500	3.150	0.0%	08-31 21:45Z
glm-5.2	SiliconFlow	siliconflow/fp8	fp8	1.190	3.740	0.0%	08-31 21:45Z
glm-5.2	StreamLake	streamlake/fp8	fp8	0.420	1.320	0.0%	08-31 21:45Z
glm-5.2	Together	together	unknown	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	Venice	venice/fp8	fp8	1.400	4.400	0.0%	08-31 21:45Z
glm-5.2	Z.AI	z-ai/fp8	fp8	1.400	4.400	0.0%	08-31 21:45Z
glm-5.3-flash	BaseTen	baseten/fp8	fp8	0.150	0.500	95.3%	08-31 21:45Z
glm-5.3-flash	Cloudflare	cloudflare	unknown	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	DeepInfra	deepinfra/fp8	fp8	0.075	0.250	0.0%	08-31 21:45Z
glm-5.3-flash	DigitalOcean	digitalocean	unknown	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Fireworks	fireworks	unknown	0.150	0.500	4.7%	08-31 21:45Z
glm-5.3-flash	Friendli	friendli	unknown	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	GMICloud	gmicloud/fp8	fp8	0.075	0.250	9.3%	08-31 21:45Z
glm-5.3-flash	Io Net	io-net/fp8	fp8	0.150	0.500	9.3%	08-31 21:45Z
glm-5.3-flash	Makora	makora	unknown	0.140	0.470	0.0%	08-31 21:45Z
glm-5.3-flash	Modal	modal/fp8	fp8	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Morph	morph/fp8	fp8	0.150	0.420	0.0%	08-31 21:45Z
glm-5.3-flash	Novita	novita/fp8	fp8	0.075	0.250	0.0%	08-31 21:45Z
glm-5.3-flash	Parasail	parasail/fp8	fp8	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Phala	phala/fp8	fp8	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Reka	reka/fp8	fp8	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Relace	relace/fp4	fp4	0.071	0.237	0.0%	08-31 21:45Z
glm-5.3-flash	SiliconFlow	siliconflow/fp8	fp8	0.150	0.500	95.3%	08-31 21:45Z
glm-5.3-flash	Together	together	unknown	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Venice	venice	unknown	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Wafer	wafer	unknown	0.150	0.500	0.0%	08-31 21:45Z
glm-5.3-flash	Z.AI	z-ai/fp8	fp8	0.075	0.250	0.0%	08-31 21:45Z
gpt-oss-120b	AkashML	akashml/bf16	bf16	0.030	0.170	0.0%	08-31 21:45Z
gpt-oss-120b	Amazon Bedrock	amazon-bedrock	unknown	0.150	0.600	14.0%	08-31 21:45Z
gpt-oss-120b	Amazon Bedrock	amazon-bedrock/eu-west-1	unknown	0.150	0.600	14.0%	08-31 21:45Z
gpt-oss-120b	BaseTen	baseten/fp4	fp4	0.100	0.500	2.3%	08-31 21:45Z
gpt-oss-120b	Cerebras	cerebras/fp16	fp16	0.350	0.750	0.0%	08-31 21:45Z
gpt-oss-120b	CoreWeave	coreweave/fp4	fp4	0.030	0.170	0.0%	08-31 21:45Z
gpt-oss-120b	DeepInfra	deepinfra/bf16	bf16	0.037	0.170	0.0%	08-31 21:45Z
gpt-oss-120b	DeepInfra	deepinfra/turbo	bf16	0.150	0.600	0.0%	08-31 21:45Z
gpt-oss-120b	DigitalOcean	digitalocean	unknown	0.055	0.385	14.0%	08-31 21:45Z
gpt-oss-120b	Google	google-vertex/global	unknown	0.090	0.360	18.6%	08-31 21:45Z
gpt-oss-120b	Groq	groq	unknown	0.150	0.600	0.0%	08-31 21:45Z
gpt-oss-120b	Mancer 2	mancer/fp8	fp8	0.065	0.500	0.0%	08-31 21:45Z
gpt-oss-120b	Mara	mara	unknown	0.150	0.750	0.0%	08-31 21:45Z
gpt-oss-120b	Nebius	nebius/fp4	fp4	0.150	0.600	0.0%	08-31 21:45Z
gpt-oss-120b	Novita	novita/fp4	fp4	0.050	0.250	0.0%	08-31 21:45Z
gpt-oss-120b	Parasail	parasail/fp4	fp4	0.100	0.750	0.0%	08-31 21:45Z
gpt-oss-120b	Phala	phala	unknown	0.150	0.600	0.0%	08-31 21:45Z
gpt-oss-120b	SambaNova	sambanova	unknown	0.140	0.950	0.0%	08-31 21:45Z
gpt-oss-120b	SiliconFlow	siliconflow/fp8	fp8	0.050	0.450	14.0%	08-31 21:45Z
gpt-oss-120b	Together	together	unknown	0.150	0.600	0.0%	08-31 21:45Z
gpt-oss-20b	Amazon Bedrock	amazon-bedrock	unknown	0.070	0.150	0.0%	08-31 21:45Z
gpt-oss-20b	Amazon Bedrock	amazon-bedrock/eu-west-1	unknown	0.070	0.150	0.0%	08-31 21:45Z
gpt-oss-20b	CoreWeave	coreweave/fp4	fp4	0.030	0.130	0.0%	08-31 21:45Z
gpt-oss-20b	Darkbloom	darkbloom/fp8	fp8	0.020	0.100	0.0%	08-31 21:45Z
gpt-oss-20b	DeepInfra	deepinfra/bf16	bf16	0.030	0.140	16.3%	08-31 21:45Z
gpt-oss-20b	Fireworks	fireworks	unknown	0.070	0.300	100.0%	08-31 21:45Z
gpt-oss-20b	Google	google-vertex/us-central1	unknown	0.070	0.250	14.0%	08-31 21:45Z
gpt-oss-20b	Groq	groq	unknown	0.075	0.300	0.0%	08-31 21:45Z
gpt-oss-20b	Novita	novita/fp4	fp4	0.040	0.150	25.6%	08-31 21:45Z
gpt-oss-20b	Parasail	parasail/fp4	fp4	0.030	0.150	0.0%	08-31 21:45Z
gpt-oss-20b	Phala	phala	unknown	0.040	0.150	30.2%	08-31 21:45Z
gpt-oss-20b	SiliconFlow	siliconflow/fp8	fp8	0.040	0.180	14.0%	08-31 21:45Z
gpt-oss-20b	Together	together	unknown	0.050	0.200	97.7%	08-31 21:45Z
hy3	AtlasCloud	atlas-cloud/fp8	fp8	0.200	0.800	0.0%	08-31 21:45Z
hy3	DeepInfra	deepinfra/fp8	fp8	0.140	0.580	0.0%	08-31 21:45Z
hy3	GMICloud	gmicloud/bf16	bf16	0.126	0.522	34.9%	08-31 21:45Z
hy3	Novita	novita	unknown	0.140	0.580	0.0%	08-31 21:45Z
hy3	Phala	phala	unknown	0.150	0.640	0.0%	08-31 21:45Z
hy3	Tencent	tencent/fp8	fp8	0.083	0.330	0.0%	08-31 21:45Z
inkling	BaseTen	baseten/fp8	fp8	1.000	4.050	30.2%	08-31 21:45Z
inkling	DeepInfra	deepinfra/fp8	fp8	0.950	4.050	0.0%	08-31 21:45Z
inkling	Together	together	unknown	1.000	4.050	0.0%	08-31 21:45Z
kimi-k3	Alibaba	alibaba	unknown	3.450	17.250	100.0%	08-31 21:45Z
kimi-k3	BaseTen	baseten/fp8	fp8	3.000	15.000	0.0%	08-31 21:45Z
kimi-k3	Chutes	chutes/mxfp4	mxfp4	3.000	15.000	14.0%	08-31 21:45Z
kimi-k3	DeepInfra	deepinfra/bf16	bf16	2.850	14.250	0.0%	08-31 21:45Z
kimi-k3	DigitalOcean	digitalocean	unknown	2.850	14.250	0.0%	08-31 21:45Z
kimi-k3	Fireworks	fireworks	unknown	3.000	15.000	4.7%	08-31 21:45Z
kimi-k3	Fireworks	fireworks/fast	unknown	4.500	22.500	18.6%	08-31 21:45Z
kimi-k3	Fireworks	fireworks/us	unknown	3.300	16.500	4.7%	08-31 21:45Z
kimi-k3	Makora	makora	unknown	2.550	12.750	2.3%	08-31 21:45Z
kimi-k3	Modal	modal/mxfp4	mxfp4	3.000	15.000	0.0%	08-31 21:45Z
kimi-k3	Moonshot AI	moonshotai/mxfp4	mxfp4	3.000	15.000	0.0%	08-31 21:45Z
kimi-k3	Morph	morph/fast	fp4	6.000	22.500	0.0%	08-31 21:45Z
kimi-k3	Morph	morph/fp4	fp4	2.800	14.000	0.0%	08-31 21:45Z
kimi-k3	Parasail	parasail/fp4	fp4	3.000	15.000	0.0%	08-31 21:45Z
kimi-k3	Phala	phala	unknown	3.000	15.000	0.0%	08-31 21:45Z
kimi-k3	Sail Research	sail-research/fp4	fp4	2.600	13.000	0.0%	08-31 21:45Z
kimi-k3	Together	together	unknown	3.000	15.000	0.0%	08-31 21:45Z
mimo-v2.5	DeepInfra	deepinfra/fp8	fp8	0.130	0.650	0.0%	08-31 21:45Z
mimo-v2.5	GMICloud	gmicloud/fp8	fp8	0.119	0.238	0.0%	08-31 21:45Z
mimo-v2.5	Novita	novita/fp8	fp8	0.168	0.336	0.0%	08-31 21:45Z
mimo-v2.5	Parasail	parasail/fp8	fp8	0.140	0.280	0.0%	08-31 21:45Z
mimo-v2.5	StreamLake	streamlake	unknown	0.168	0.336	14.0%	08-31 21:45Z
mimo-v2.5	Xiaomi	xiaomi/fp8	fp8	0.140	0.280	0.0%	08-31 21:45Z
mimo-v2.5-pro	AtlasCloud	atlas-cloud/fp8	fp8	0.435	0.870	0.0%	08-31 21:45Z
mimo-v2.5-pro	DeepInfra	deepinfra/fp8	fp8	0.390	1.170	0.0%	08-31 21:45Z
mimo-v2.5-pro	DigitalOcean	digitalocean	unknown	0.400	1.500	0.0%	08-31 21:45Z
mimo-v2.5-pro	GMICloud	gmicloud/bf16	bf16	0.304	0.609	16.3%	08-31 21:45Z
mimo-v2.5-pro	Novita	novita	unknown	0.480	0.961	0.0%	08-31 21:45Z
mimo-v2.5-pro	StreamLake	streamlake	unknown	0.522	1.044	0.0%	08-31 21:45Z
mimo-v2.5-pro	Xiaomi	xiaomi/fp8	fp8	0.435	0.870	0.0%	08-31 21:45Z
minimax-m3	AtlasCloud	atlas-cloud/fp8	fp8	0.300	1.200	0.0%	08-31 21:45Z
minimax-m3	CoreWeave	coreweave/fp4	fp4	0.230	0.960	14.0%	08-31 21:45Z
minimax-m3	DeepInfra	deepinfra/fp8	fp8	0.280	1.100	0.0%	08-31 21:45Z
minimax-m3	GMICloud	gmicloud/fp8	fp8	0.600	2.400	14.0%	08-31 21:45Z
minimax-m3	Minimax	minimax/fp8	fp8	0.300	1.200	0.0%	08-31 21:45Z
minimax-m3	ModelRun	modelrun/fp4	fp4	0.750	3.000	0.0%	08-31 21:45Z
minimax-m3	Novita	novita/fp8	fp8	0.300	1.200	0.0%	08-31 21:45Z
minimax-m3	Parasail	parasail/fp8	fp8	0.300	1.200	0.0%	08-31 21:45Z
minimax-m3	SambaNova	sambanova	unknown	0.600	2.400	14.0%	08-31 21:45Z
minimax-m3	StreamLake	streamlake/fp8	fp8	0.300	1.200	14.0%	08-31 21:45Z
minimax-m3	Venice	venice/fp8	fp8	0.300	1.200	0.0%	08-31 21:45Z
muse-glimmer-30b	DeepInfra	deepinfra/bf16	bf16	0.300	1.200	0.0%	08-31 21:45Z
muse-glimmer-30b	Fireworks	fireworks	unknown	0.350	1.500	4.7%	08-31 21:45Z
muse-glimmer-30b	Parasail	parasail/bf16	bf16	0.300	1.100	0.0%	08-31 21:45Z
muse-glimmer-30b	Phala	phala	unknown	0.300	1.100	0.0%	08-31 21:45Z
muse-glimmer-30b	Together	together	unknown	0.350	1.500	0.0%	08-31 21:45Z
nemotron3-ultra	BaseTen	baseten/fp4	fp4	0.600	2.400	7.0%	08-31 21:45Z
nemotron3-ultra	DeepInfra	deepinfra/fp4	fp4	0.500	2.200	100.0%	08-31 21:45Z
nemotron3-ultra	Venice	venice/fp8	fp8	0.625	3.125	0.0%	08-31 21:45Z
qwen36-35b-a3b	AkashML	akashml/fp8	fp8	0.100	0.900	0.0%	08-31 21:45Z
qwen36-35b-a3b	AtlasCloud	atlas-cloud/fp8	fp8	0.186	1.114	0.0%	08-31 21:45Z
qwen36-35b-a3b	CoreWeave	coreweave/fp8	fp8	0.250	1.250	14.0%	08-31 21:45Z
qwen36-35b-a3b	Darkbloom	darkbloom/fp4	fp4	0.050	0.700	2.3%	08-31 21:45Z
qwen36-35b-a3b	DeepInfra	deepinfra/fp8	fp8	0.100	0.950	0.0%	08-31 21:45Z
qwen36-35b-a3b	Io Net	io-net/fp8	fp8	0.190	1.190	0.0%	08-31 21:45Z
qwen36-35b-a3b	Parasail	parasail/fp8	fp8	0.150	1.000	0.0%	08-31 21:45Z
qwen36-35b-a3b	Phala	phala	unknown	0.200	1.270	0.0%	08-31 21:45Z
qwen36-35b-a3b	SiliconFlow	siliconflow/fp8	fp8	0.200	1.600	14.0%	08-31 21:45Z
qwen36-35b-a3b	Venice	venice/fp8	fp8	0.100	1.000	0.0%	08-31 21:45Z

An en dash means we do not have a number we trust for that cell, and we print nothing rather than a guess. 238 endpoints are tracked across the full 12-day window; the table above is the 219 seen on the most recent sweep day.

Everything that moved

The home page shows the biggest moves. This is the whole list for the window 2026-08-20 to 2026-08-31 — every move that cleared our published margin, nothing curated out.

Every price move this window

All 91 price moves that cleared the margin, newest first, dollars per million tokens. Iris rises, teal falls; a 0% change on one side means the other side is what moved.

model	provider	pin	$/M in, before → after	change	$/M out, before → after	change	caught
deepseek-v4-pro	StreamLake	streamlake/fp8	0.417 → 1.028	+146%	0.835 → 2.055	+146%	2026-08-31
glm-5.2	Novita	novita/fp8	0.342 → 0.742	+117%	1.074 → 2.332	+117%	2026-08-31
deepseek-v4-pro	Baidu	baidu/fp8	0.426 → 0.869	+104%	0.852 → 1.737	+104%	2026-08-31
glm-5.3-flash	Relace	relace/fp4	0.050 → 0.071	+42%	0.167 → 0.237	+42%	2026-08-31
glm-5.2	StreamLake	streamlake/fp8	0.328 → 0.420	+28%	1.030 → 1.320	+28%	2026-08-31
glm-5.2	Baidu	baidu/fp8	0.372 → 0.419	+12%	1.170 → 1.316	+12%	2026-08-31
gpt-oss-120b	Mancer 2	mancer/fp8	0.070 → 0.065	−7%	0.500 → 0.500	0%	2026-08-31
glm-5.3-flash	Makora	makora	0.150 → 0.140	−7%	0.500 → 0.470	−6%	2026-08-31
deepseek-v4-flash	Mancer 2	mancer/fp8	0.170 → 0.180	+6%	0.500 → 0.500	0%	2026-08-31
deepseek-v4-flash	StreamLake	streamlake/fp8	0.079 → 0.081	+3%	0.157 → 0.162	+3%	2026-08-31
deepseek-v4-flash	Baidu	baidu/fp8	0.079 → 0.081	+2%	0.158 → 0.162	+2%	2026-08-31
glm-5.3-flash	Relace	relace/fp4	0.075 → 0.050	−33%	0.250 → 0.167	−33%	2026-08-30
deepseek-v4-pro	StreamLake	streamlake/fp8	0.539 → 0.417	−23%	1.078 → 0.835	−23%	2026-08-30
deepseek-v4-pro	Baidu	baidu/fp8	0.548 → 0.426	−22%	1.095 → 0.852	−22%	2026-08-30
gpt-oss-120b	Mancer 2	mancer/fp8	0.085 → 0.070	−18%	0.500 → 0.500	0%	2026-08-30
glm-5.3-flash	Morph	morph/fp8	0.150 → 0.150	0%	0.500 → 0.420	−16%	2026-08-30
deepseek-v4-flash	Baidu	baidu/fp8	0.082 → 0.079	−4%	0.164 → 0.158	−4%	2026-08-30
deepseek-v4-flash	StreamLake	streamlake/fp8	0.082 → 0.079	−4%	0.164 → 0.157	−4%	2026-08-30
deepseek-v4-flash	Mancer 2	mancer/fp8	0.165 → 0.170	+3%	0.500 → 0.500	0%	2026-08-30
glm-5.2	Novita	novita/fp8	0.742 → 0.342	−54%	2.332 → 1.074	−54%	2026-08-29
deepseek-v4-flash	DigitalOcean	digitalocean	0.140 → 0.068	−52%	0.280 → 0.168	−40%	2026-08-29
deepseek-v3.2	DigitalOcean	digitalocean	0.500 → 0.250	−50%	1.600 → 0.800	−50%	2026-08-29
deepseek-v4-pro	DigitalOcean	digitalocean	1.740 → 0.870	−50%	3.480 → 1.740	−50%	2026-08-29
glm-5.2	DigitalOcean	digitalocean	1.400 → 0.700	−50%	4.400 → 2.200	−50%	2026-08-29
glm-5.3-flash	DeepInfra	deepinfra/fp8	0.150 → 0.075	−50%	0.500 → 0.250	−50%	2026-08-29
mimo-v2.5-pro	DigitalOcean	digitalocean	0.800 → 0.400	−50%	3.000 → 1.500	−50%	2026-08-29
gpt-oss-120b	DigitalOcean	digitalocean	0.100 → 0.055	−45%	0.700 → 0.385	−45%	2026-08-29
deepseek-v4-pro	GMICloud	gmicloud/fp8	0.792 → 1.044	+32%	2.376 → 2.088	−12%	2026-08-29
deepseek-v4-pro	StreamLake	streamlake/fp8	0.742 → 0.539	−27%	1.483 → 1.078	−27%	2026-08-29
deepseek-v4-pro	Baidu	baidu/fp8	0.750 → 0.548	−27%	1.501 → 1.095	−27%	2026-08-29
glm-5.2	StreamLake	streamlake/fp8	0.420 → 0.328	−22%	1.320 → 1.030	−22%	2026-08-29
gemma4-31b	OpenInference	open-inference/bf16	0.080 → 0.070	−12%	0.350 → 0.350	0%	2026-08-29
deepseek-v4-flash	Mancer 2	mancer/fp8	0.160 → 0.165	+3%	0.450 → 0.500	+11%	2026-08-29
glm-5.2	Baidu	baidu/fp8	0.419 → 0.372	−11%	1.316 → 1.170	−11%	2026-08-29
deepseek-v4-flash	Baidu	baidu/fp8	0.087 → 0.082	−6%	0.174 → 0.164	−6%	2026-08-29
deepseek-v4-flash	StreamLake	streamlake/fp8	0.087 → 0.082	−6%	0.174 → 0.164	−6%	2026-08-29
kimi-k3	DigitalOcean	digitalocean	3.000 → 2.850	−5%	15.000 → 14.250	−5%	2026-08-29
deepseek-v4-flash	DigitalOcean	digitalocean	0.068 → 0.140	+106%	0.168 → 0.280	+67%	2026-08-28
deepseek-v3.2	DigitalOcean	digitalocean	0.250 → 0.500	+100%	0.800 → 1.600	+100%	2026-08-28
deepseek-v4-pro	DigitalOcean	digitalocean	0.870 → 1.740	+100%	1.740 → 3.480	+100%	2026-08-28
glm-5.2	DigitalOcean	digitalocean	0.700 → 1.400	+100%	2.200 → 4.400	+100%	2026-08-28
mimo-v2.5-pro	DigitalOcean	digitalocean	0.400 → 0.800	+100%	1.500 → 3.000	+100%	2026-08-28
gpt-oss-120b	DigitalOcean	digitalocean	0.055 → 0.100	+82%	0.385 → 0.700	+82%	2026-08-28
hy3	Baidu	baidu/fp8	0.082 → 0.132	+60%	0.330 → 0.527	+60%	2026-08-28
gemma4-31b	Together	together	0.280 → 0.390	+39%	0.860 → 0.970	+13%	2026-08-28
glm-5.2	DeepInfra	deepinfra/fp4	0.750 → 0.487	−35%	2.400 → 1.560	−35%	2026-08-28
deepseek-v4-pro	StreamLake	streamlake/fp8	0.870 → 0.742	−15%	1.740 → 1.483	−15%	2026-08-28
deepseek-v4-flash	StreamLake	streamlake/fp8	0.076 → 0.087	+14%	0.153 → 0.174	+14%	2026-08-28
deepseek-v4-flash	Baidu	baidu/fp8	0.078 → 0.087	+12%	0.156 → 0.174	+12%	2026-08-28
deepseek-v4-flash	Mancer 2	mancer/fp8	0.150 → 0.160	+7%	0.450 → 0.450	0%	2026-08-28
gpt-oss-120b	Mancer 2	mancer/fp8	0.080 → 0.085	+6%	0.500 → 0.500	0%	2026-08-28
glm-5.2	Inceptron	inceptron/fp4	0.750 → 0.710	−5%	2.400 → 2.350	−2%	2026-08-28
kimi-k3	DigitalOcean	digitalocean	2.850 → 3.000	+5%	14.250 → 15.000	+5%	2026-08-28
deepseek-v4-pro	Baidu	baidu/fp8	0.791 → 0.750	−5%	1.582 → 1.501	−5%	2026-08-28
glm-5.2	Novita	novita/fp8	0.409 → 0.742	+82%	1.285 → 2.332	+82%	2026-08-27
qwen36-35b-a3b	AkashML	akashml/fp8	0.140 → 0.100	−29%	1.000 → 0.900	−10%	2026-08-27
qwen36-35b-a3b	Darkbloom	darkbloom/fp4	0.070 → 0.050	−29%	0.700 → 0.700	0%	2026-08-27
gpt-oss-120b	Mancer 2	mancer/fp8	0.085 → 0.080	−6%	0.500 → 0.500	0%	2026-08-27
deepseek-v4-flash	StreamLake	streamlake/fp8	0.080 → 0.076	−4%	0.159 → 0.153	−4%	2026-08-27
deepseek-v4-flash	Baidu	baidu/fp8	0.081 → 0.078	−4%	0.162 → 0.156	−4%	2026-08-27
deepseek-v4-flash	Mancer 2	mancer/fp8	0.155 → 0.150	−3%	0.450 → 0.450	0%	2026-08-27
glm-5.2	Baidu	baidu/fp8	0.410 → 0.419	+2%	1.289 → 1.316	+2%	2026-08-27
mimo-v2.5-pro	DeepInfra	deepinfra/fp8	1.000 → 0.390	−61%	3.000 → 1.170	−61%	2026-08-26
deepseek-v4-pro	StreamLake	streamlake/fp8	0.556 → 0.870	+56%	1.112 → 1.740	+56%	2026-08-26
deepseek-v4-pro	Baidu	baidu/fp8	0.558 → 0.791	+42%	1.115 → 1.582	+42%	2026-08-26
hy3	Tencent	tencent/fp8	0.132 → 0.083	−38%	0.528 → 0.330	−38%	2026-08-26
hy3	Baidu	baidu/fp8	0.125 → 0.082	−34%	0.502 → 0.330	−34%	2026-08-26
glm-5.2	Io Net	io-net/fp4	0.990 → 0.680	−31%	3.600 → 3.300	−8%	2026-08-26
glm-5.2	StreamLake	streamlake/fp8	0.560 → 0.420	−25%	1.760 → 1.320	−25%	2026-08-26
glm-5.2	Baidu	baidu/fp8	0.462 → 0.410	−11%	1.452 → 1.289	−11%	2026-08-26
glm-5.2	Novita	novita/fp8	0.461 → 0.409	−11%	1.448 → 1.285	−11%	2026-08-26
deepseek-v4-flash	Mancer 2	mancer/fp8	0.140 → 0.155	+11%	0.450 → 0.450	0%	2026-08-26
deepseek-v4-flash	StreamLake	streamlake/fp8	0.080 → 0.080	−0%	0.160 → 0.159	−0%	2026-08-26
deepseek-v4-flash	Baidu	baidu/fp8	0.081 → 0.081	−0%	0.162 → 0.162	−0%	2026-08-26
gpt-oss-120b	AkashML	akashml/bf16	0.037 → 0.030	−19%	0.490 → 0.170	−65%	2026-08-25
deepseek-v4-flash	Baidu	baidu/fp8	0.055 → 0.081	+49%	0.109 → 0.162	+49%	2026-08-25
deepseek-v4-flash	StreamLake	streamlake/fp8	0.056 → 0.080	+42%	0.112 → 0.160	+42%	2026-08-25
glm-5.2	Novita	novita/fp8	0.700 → 0.461	−34%	2.200 → 1.448	−34%	2026-08-25
glm-5.2	StreamLake	streamlake/fp8	0.693 → 0.560	−19%	2.178 → 1.760	−19%	2026-08-25
glm-5.2	Inceptron	inceptron/fp4	0.750 → 0.750	0%	2.900 → 2.400	−17%	2026-08-25
deepseek-v4-pro	Baidu	baidu/fp8	0.507 → 0.558	+10%	1.014 → 1.115	+10%	2026-08-25
deepseek-v4-pro	StreamLake	streamlake/fp8	0.522 → 0.556	+6%	1.044 → 1.112	+6%	2026-08-25
glm-5.2	Baidu	baidu/fp8	0.490 → 0.462	−6%	1.540 → 1.452	−6%	2026-08-25
glm-5.2	StreamLake	streamlake/fp8	0.336 → 0.693	+106%	1.056 → 2.178	+106%	2026-08-24
deepseek-v4-flash	GMICloud	gmicloud/fp8	0.059 → 0.112	+90%	0.118 → 0.224	+90%	2026-08-24
glm-5.2	Baidu	baidu/fp8	0.350 → 0.490	+40%	1.100 → 1.540	+40%	2026-08-24
glm-5.2	GMICloud	gmicloud/fp8	1.400 → 1.050	−25%	4.400 → 3.300	−25%	2026-08-24
deepseek-v4-flash	StreamLake	streamlake/fp8	0.049 → 0.056	+15%	0.098 → 0.112	+15%	2026-08-24
deepseek-v4-flash	Baidu	baidu/fp8	0.049 → 0.055	+11%	0.098 → 0.109	+11%	2026-08-24
qwen36-35b-a3b	Darkbloom	darkbloom/fp4	0.080 → 0.070	−12%	0.750 → 0.700	−7%	2026-08-23
gpt-oss-120b	Mancer 2	mancer/fp8	0.080 → 0.085	+6%	0.500 → 0.500	0%	2026-08-23

Every row is one adjudicated day-over-day change on one endpoint, caught by the sweep dated in the last column.

Every latency move this window

All 30 median response-time moves that cleared the margin, newest first, milliseconds. Iris slower, teal faster.

model	provider	pin	median ms, before → after	change	caught
gpt-oss-120b	Novita	novita/fp4	1,636 → 27,620	+1588%	2026-08-31
glm-5.2	Ambient	ambient/fp8	2,803 → 12,575	+349%	2026-08-31
kimi-k3	Together	together	1,570 → 4,548	+190%	2026-08-31
inkling	DeepInfra	deepinfra/fp8	94,078 → 3,072	−97%	2026-08-31
deepseek-v4-pro	DigitalOcean	digitalocean	88,359 → 15,092	−83%	2026-08-31
mimo-v2.5	Xiaomi	xiaomi/fp8	20,019 → 6,976	−65%	2026-08-31
mimo-v2.5	Xiaomi	xiaomi/fp8	6,001 → 20,019	+234%	2026-08-30
kimi-k3	Morph	morph/fast	10,593 → 23,692	+124%	2026-08-30
inkling	Together	together	39,362 → 1,780	−96%	2026-08-30
muse-glimmer-30b	DeepInfra	deepinfra/bf16	49,674 → 2,322	−95%	2026-08-30
inkling	DeepInfra	deepinfra/fp8	2,537 → 59,975	+2264%	2026-08-29
inkling	Together	together	2,522 → 39,362	+1461%	2026-08-29
kimi-k3	DeepInfra	deepinfra/bf16	3,042 → 10,074	+231%	2026-08-29
kimi-k3	Sail Research	sail-research/fp4	9,067 → 1,741	−81%	2026-08-29
gpt-oss-120b	SiliconFlow	siliconflow/fp8	20,321 → 5,421	−73%	2026-08-29
muse-glimmer-30b	DeepInfra	deepinfra/bf16	1,632 → 28,535	+1648%	2026-08-28
qwen36-35b-a3b	SiliconFlow	siliconflow/fp8	13,940 → 35,410	+154%	2026-08-28
kimi-k3	DeepInfra	deepinfra/bf16	14,735 → 3,042	−79%	2026-08-28
kimi-k3	Together	together	6,675 → 1,929	−71%	2026-08-28
mimo-v2.5	Xiaomi	xiaomi/fp8	10,209 → 4,243	−58%	2026-08-28
nemotron3-ultra	Venice	venice/fp8	3,136 → 24,381	+678%	2026-08-27
muse-glimmer-30b	DeepInfra	deepinfra/bf16	12,030 → 1,632	−86%	2026-08-27
muse-glimmer-30b	DeepInfra	deepinfra/bf16	1,638 → 12,030	+634%	2026-08-26
deepseek-v4-pro	Ionstream	ionstream/fp4	33,139 → 7,743	−77%	2026-08-26
gpt-oss-120b	SiliconFlow	siliconflow/fp8	6,349 → 27,950	+340%	2026-08-25
gemma4-31b	Novita	novita/bf16	22,276 → 5,731	−74%	2026-08-25
glm-5.2	Venice	venice/fp8	6,058 → 1,952	−68%	2026-08-25
glm-5.2	Venice	venice/fp8	1,745 → 6,058	+247%	2026-08-24
kimi-k3	Morph	morph/fast	6,509 → 12,783	+96%	2026-08-24
deepseek-v4-flash	Phala	phala	30,672 → 6,305	−79%	2026-08-24

A latency change is only reported once the two measurement windows stop overlapping, so noise does not become a move.

The grid, live.

The endpoint grid

Everything that moved

Every price move this window

Every latency move this window