BroadRouter

Llama 3 70B vs Llama 3 8B API pricing

Updated

Llama 3 70B: USD 2.65 input, USD 3.50 output per 1M tokens (Amazon Bedrock). Llama 3 8B: USD 0.0925 input, USD 0.095 output per 1M tokens (Featherless). At 3 input tokens per output token, Llama 3 70B costs 31× as much as Llama 3 8B (USD 2.862 vs USD 0.09312 per 1M).

Llama 3 70B and Llama 3 8B are neighbouring sizes of the same Meta generation (70B and 8B parameters, by name). Price only: every number is computed from prices read from each provider's own pricing page (linked per cell, newest check 2026-11-15), with no claim about quality, speed or accuracy.

How do Llama 3 70B and Llama 3 8B differ on price?

Llama 3 70B and Llama 3 8B prices side by side

USD per 1M tokens, standard tier. "Cheapest" is the lowest price at 3 input tokens to 1 output token; each line can come from a different provider.

Llama 3 70B and Llama 3 8B API prices compared
PriceLlama 3 70BLlama 3 8B
Model pageLlama 3 Instruct (70B)
Meta · 2 providers
Llama 3 8B Instruct
Meta · 3 providers
Cheapest standard price, blended 3:1USD 2.862
USD 2.65 in / USD 3.50 out at Amazon Bedrock
Amazon Bedrock page checked 2026-11-14
USD 0.09312 lower
USD 0.0925 in / USD 0.095 out at Featherless
Featherless page checked 2026-11-15
Cheapest inputUSD 2.65
Amazon Bedrock source
USD 0.0925 lower
Featherless source
Cheapest cached inputUSD 0.70
Featherless source
USD 0.074 lower
Featherless source
Cheapest outputUSD 3.50
Amazon Bedrock source
USD 0.095 lower
Featherless source
Maker's own APIMeta publishes no paid standard price on its own API that we found.Meta publishes no paid standard price on its own API that we found.
Median standard priceUSD 3.075 / USD 5.90
input / output, 2 prices
USD 0.282 / USD 0.60
input / output, 3 prices
Context window8,192 tokens, largest stated8,192 tokens, largest stated
Providers with a standard price23

Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).

What do Llama 3 70B and Llama 3 8B cost for a real workload?

30 days at each model's cheapest standard price for the workload (presets of the cost calculator; cached input at the cached price where listed).

Monthly cost of three workloads on Llama 3 70B and Llama 3 8B
WorkloadLlama 3 70BLlama 3 8BDifference
Support chatbot
2,000 requests a day, 1,500 input tokens (60% cached), 300 output
USD 301.50
Amazon Bedrock
USD 9.04
Featherless
USD 292.46
Llama 3 8B lower
RAG over documents
1,000 requests a day, 8,000 input tokens (30% cached), 500 output
USD 688.50
Amazon Bedrock
USD 80.09
Cloudflare Workers AI
USD 608.42
Llama 3 8B lower
Coding agent (one developer)
1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning
USD 5,085
Amazon Bedrock
USD 582.03
Cloudflare Workers AI
USD 4,503
Llama 3 8B lower

Llama 3 8B is cheaper on all three workloads. The widest gap is on Support chatbot: the dearer model costs 33× as much as the cheaper one.

Does the cheaper model change with your input:output ratio?

Llama 3 8B is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below Llama 3 70B's, so the input:output ratio never changes the answer.

Price per 1M tokens of Llama 3 70B and Llama 3 8B at different input:output mixes
Input : output tokensLlama 3 70BLlama 3 8B
1 : 3 (long answers)USD 3.287
Amazon Bedrock
USD 0.09438 lower
Featherless
1 : 1USD 3.075
Amazon Bedrock
USD 0.09375 lower
Featherless
3 : 1 (the site’s blend)USD 2.862
Amazon Bedrock
USD 0.09312 lower
Featherless
10 : 1 (retrieval, long prompts)USD 2.727
Amazon Bedrock
USD 0.09273 lower
Featherless
30 : 1 (agents, big contexts)USD 2.677
Amazon Bedrock
USD 0.09258 lower
Featherless

USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.

What do Llama 3 70B and Llama 3 8B cost per character in other languages?

Cost to send 1M characters at each model's cheapest non-promotional standard input price, with the Llama 3.3 70B tokenizer for Llama 3 70B (the same one for both, so the gap equals the price gap):

Input cost per 1M characters by language, Llama 3 70B and Llama 3 8B
ContentLlama 3 70BLlama 3 8B
EnglishUSD 0.575
Amazon Bedrock
USD 0.020 lower
Featherless
Arabic MSAUSD 1.02
Amazon Bedrock
USD 0.036 lower
Featherless
Egyptian ArabicUSD 1.04
Amazon Bedrock
USD 0.036 lower
Featherless
HindiUSD 1.38
Amazon Bedrock
USD 0.048 lower
Featherless
ChineseUSD 2.12
Amazon Bedrock
USD 0.074 lower
Featherless
SpanishUSD 0.708
Amazon Bedrock
USD 0.025 lower
Featherless
CodeUSD 0.674
Amazon Bedrock
USD 0.024 lower
Featherless

From the real cost by language dataset (measured tokens per character on FLORES-200 and other texts). Your own text will differ.

Check it on your own traffic

More comparisons: Llama 3.1 70B vs Llama 3 70B · Llama 3.1 8B Instruct vs Llama 3 8B · every Llama 3 70B comparison · every Llama 3 8B comparison · all 662 price comparisons

Questions

Is Llama 3 70B cheaper than Llama 3 8B?

No. On a blend of 3 input tokens to 1 output token, Llama 3 8B costs USD 0.09312 per 1M tokens and Llama 3 70B USD 2.862, at each model's cheapest standard price. For input alone, Llama 3 8B is cheaper (USD 0.0925 against USD 2.65 per 1M); for output alone, Llama 3 8B is (USD 0.095 against USD 3.50). This compares price only, not quality.

What do Llama 3 70B and Llama 3 8B cost per 1M tokens?

Llama 3 70B: from USD 2.65 input and USD 3.50 output at Amazon Bedrock; median of 2 standard prices USD 3.075 / USD 5.90. Llama 3 8B: from USD 0.0925 input and USD 0.095 output at Featherless; median of 3 standard prices USD 0.282 / USD 0.60. Each price is read from the provider's own pricing page; the source is linked in the table.

Which providers sell both Llama 3 70B and Llama 3 8B?

2 providers have a standard price for both: Amazon Bedrock (USD 2.862 and USD 0.375 at 3:1), Featherless (USD 4.70 and USD 0.09312 at 3:1). The first is the lowest combined price. Buying each model from its own cheapest provider costs USD 2.862 and USD 0.09312.

Which is cheaper for a coding agent or a chatbot, Llama 3 70B or Llama 3 8B?

For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 5,085 a month on Llama 3 70B (Amazon Bedrock) against USD 582.03 on Llama 3 8B (Cloudflare Workers AI), a difference of USD 4,503. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 301.50 a month on Llama 3 70B (Amazon Bedrock) against USD 9.04 on Llama 3 8B (Featherless), a difference of USD 292.46. Each uses the cheapest standard price for that workload, cached input included. Try your own numbers.

At what input:output ratio does the cheaper of Llama 3 70B and Llama 3 8B change?

Llama 3 8B is cheaper at every mix of input and output tokens, before cache discounts: its cheapest input price and its cheapest output price are both at or below Llama 3 70B's, so the input:output ratio never changes the answer.

Do Llama 3 70B and Llama 3 8B have cached input pricing?

Llama 3 70B from USD 0.70 per 1M cached input tokens (Featherless; 1 of 2 standard prices list a discount). Llama 3 8B from USD 0.074 per 1M cached input tokens (Featherless; 1 of 3 standard prices list a discount). Cached tokens are billed at the input price where no cache price is published.

Which has the larger context window, Llama 3 70B or Llama 3 8B?

Both go up to 8,192 tokens on the providers that state it.

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity