Mixtral 8x7B API pricing

4 providers serve Mixtral 8x7B. Mixtral 8x7B is an open-weight model (Apache-2.0) with 46.7B total parameters, 12.9B active per token (mixture-of-experts), up to 33K context. Prices below are per 1M tokens, cheapest input first (self-hosting math further down).

Refreshed 2026-08-23 · list prices from provider APIs

Input floor $0.070/1M at Perplexity Output floor $0.280/1M at Perplexity Find your workload winner →

Provider	Input $/1M	Output $/1M	Cache read $/1M	Context
Perplexity	$0.070	$0.280output floor	-	4K	Use
AWS Bedrock	$0.450	$0.700	-	32K	Use
Databricks	$0.500	$1.00	-	4K	Use
Fireworks AI	$0.500	$0.500	-	33K	Use

The cheapest way to run Mixtral 8x7B

On an illustrative 70M-input / 30M-output monthly workload, today's lowest listed cost is roughly $13.30/month, at Perplexity. Input-heavy, output-heavy and cache-heavy workloads can produce different winners. Use the live mix calculator below rather than combining floors from two different providers. On output-token cost alone, a busy self-hosted deployment can work out cheaper (~$0.128/1M output on A100 PCIe). See the math below. Run your own numbers in the breakeven calculator, or read the breakeven math.

Your actual monthly cost

Uses today's list prices

Workload preset Input tokens / month (millions) Output tokens / month (millions) Input served from cache %

Cache read uses the listed cache rate where available; otherwise it falls back to normal input price. Batch, write-cache, volume and negotiated discounts are excluded.

Or run it yourself: Mixtral 8x7B self-hosting economics

Mixtral 8x7B is open-weight (Apache-2.0), so the API price above competes with the GPU-hour market. 46.7B total parameters (MoE, ~12.9B active per token) needs roughly 65 GB VRAM at FP8 or 33 GB at INT4, KV-cache headroom included.

GPU that fits (FP8, single card)	VRAM	From $/hr	Cheapest at
A100 PCIe	80 GB	$0.442marketplace	Vast.ai	Rent
RTX PRO 6000	96 GB	$0.500secure	RunPod	Rent
H200 NVL	141 GB	$0.500community	RunPod	Rent
MI300X	192 GB	$0.500community	RunPod	Rent
A100 SXM 80GB	80 GB	$0.667marketplace	Vast.ai	Rent
H100 SXM	80 GB	$1.34marketplace	Vast.ai	Rent

Breakeven estimate: a well-batched vLLM deployment on one A100 PCIe ($0.442/hr) at ~1918 aggregate tok/s and 50% utilization works out to roughly $0.128 per 1M output tokens, versus $0.280 via the cheapest output-token API (Perplexity). Self-hosting wins on cost if you can keep the GPU busy. Planning estimate: throughput varies with hardware, quantization, batch size and context. Tune it in the calculator.

Related models

Model	Cheapest in $/1M	Cheapest out $/1M	Providers
Mixtral-8x7B-Instruct-V0.1	$0.150	$0.150	5

FAQ

What is the cheapest API for Mixtral 8x7B?

As of 2026-08-23, the lowest input price for Mixtral 8x7B is Perplexity at $0.070 per 1M input tokens. The lowest output price is Perplexity at $0.280 per 1M output tokens. The cheapest provider for a real workload depends on its input/output mix. The next-lowest input price is AWS Bedrock at $0.450, a 543% difference.

How much VRAM do you need to self-host Mixtral 8x7B?

Mixtral 8x7B has 46.7B parameters, so plan for roughly 65 GB of VRAM at FP8 or 33 GB at INT4/AWQ, KV-cache headroom included. A single A100 PCIe (80 GB) fits it.

← All models · GPU rental prices · Breakeven calculator · Get this data as JSON