gpt-oss-20b vs Llama 3.1 8B Instruct API pricing
Updated
gpt-oss-20b: USD 0.018 input, USD 0.09 output per 1M tokens (AkashML, promotional). Llama 3.1 8B Instruct: USD 0.02 input, USD 0.04 output per 1M tokens (DeepInfra). At 3 input tokens per output token, gpt-oss-20b costs 44% more than Llama 3.1 8B Instruct (USD 0.036 vs USD 0.025 per 1M).
How do gpt-oss-20b and Llama 3.1 8B Instruct differ on price?
- Cheapest input price: gpt-oss-20b USD 0.018 against USD 0.02 per 1M tokens, so Llama 3.1 8B Instruct costs 1.11× as much (11% more).
- Cheapest output price: Llama 3.1 8B Instruct USD 0.04 against USD 0.09 per 1M tokens, so gpt-oss-20b costs 2.3× as much.
- Among the 201 indexed models, gpt-oss-20b ranks 3rd cheapest and Llama 3.1 8B Instruct 1st on the 3:1 blend.
- The cheaper model changes at 25 input tokens per output token.
- Both list cached-input discounts: gpt-oss-20b at 4 standard prices of 10, from USD 0.007 (Google Vertex AI, 90% below that provider's input price); Llama 3.1 8B Instruct at 1 standard price of 6, from USD 0.028 (Featherless, 86% below that provider's input price).
- Both state a context window of up to 131,072 tokens.
- gpt-oss-20b has 10 providers with a standard price, 4 more than Llama 3.1 8B Instruct; 5 providers sell both (DeepInfra, CoreWeave Forge, Featherless, Amazon Bedrock and 1 more).
- On one account, DeepInfra charges USD 0.0575 for gpt-oss-20b (1.6× its cheapest price) and USD 0.025 for Llama 3.1 8B Instruct (its cheapest price) at 3:1, the lowest combined price among the providers that sell both.
- Shopping around matters more for Llama 3.1 8B Instruct: its dearest standard price is 17× its cheapest, against 6.3× for gpt-oss-20b.
- gpt-oss-20b has a batch price at 2 providers, from USD 0.0575 at 3:1 (Google Vertex AI); Llama 3.1 8B Instruct has a batch price at 1 provider, from USD 0.11 at 3:1 (Amazon Bedrock).
- For the one-developer coding agent preset, the gap is USD 154.44 a year in gpt-oss-20b's favour (gpt-oss-20b USD 26.73 a month, Llama 3.1 8B Instruct USD 39.60).
- Per character of input, the gap depends on the language: gpt-oss-20b costs 1.5× as much as Llama 3.1 8B Instruct on Code text but 7% less than Llama 3.1 8B Instruct on Hindi, because the two tokenizers split those languages differently.
gpt-oss-20b and Llama 3.1 8B Instruct 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.
| Price | gpt-oss-20b | Llama 3.1 8B Instruct |
|---|---|---|
| Model page | GPT OSS 20B OpenAI · 10 providers | Llama 3.1 8B Instruct Meta · 6 providers |
| Cheapest standard price, blended 3:1 | USD 0.036 promo USD 0.018 in / USD 0.09 out at AkashML AkashML page checked 2026-11-15 | USD 0.025 lower USD 0.02 in / USD 0.04 out at DeepInfra DeepInfra page checked 2026-11-14 |
| Cheapest input | USD 0.018 lower promo AkashML source | USD 0.02 DeepInfra source |
| Cheapest cached input | USD 0.007 lower Google Vertex AI source | USD 0.028 Featherless source |
| Cheapest output | USD 0.09 promo AkashML source | USD 0.04 lower DeepInfra source |
| Cheapest non-promotional price | USD 0.03 / USD 0.13 CoreWeave Forge, USD 0.055 blended source | The cheapest price is not promotional. |
| Maker's own API | OpenAI 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 price | USD 0.04 / USD 0.19 input / output, 10 prices | USD 0.21 / USD 0.22 input / output, 6 prices |
| Context window | 131,072 tokens, largest stated | 131,072 tokens, largest stated; from 16,384 at some providers |
| Providers with a standard price | 10 | 6 |
- Lower marks the lower price on a line; tie, equal prices. Other tiers and free rows are on the model pages.
- Promo: the provider calls the price promotional or discounted, so it is current now and can end without notice.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What do gpt-oss-20b and Llama 3.1 8B Instruct 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).
| Workload | gpt-oss-20b | Llama 3.1 8B Instruct | Difference |
|---|---|---|---|
| Support chatbot 2,000 requests a day, 1,500 input tokens (60% cached), 300 output | USD 2.75 AkashML | USD 2.52 DeepInfra | USD 0.23 Llama 3.1 8B Instruct lower |
| RAG over documents 1,000 requests a day, 8,000 input tokens (30% cached), 500 output | USD 5.02 AkashML | USD 5.40 DeepInfra | USD 0.38 gpt-oss-20b lower |
| Coding agent (one developer) 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning | USD 26.73 AkashML | USD 39.60 DeepInfra | USD 12.87 gpt-oss-20b lower |
The cheaper model depends on the workload: Llama 3.1 8B Instruct for Support chatbot, gpt-oss-20b for RAG over documents, gpt-oss-20b for Coding agent (one developer). Cache-heavy work rewards a low cached-input price. The widest gap is on Coding agent (one developer): the dearer model costs 48% more than the cheaper one.
Does the cheaper model change with your input:output ratio?
The cheaper model changes with the mix: below 25 input tokens per output token, Llama 3.1 8B Instruct is cheaper; above 25, gpt-oss-20b is cheaper. At the site's 3:1 blend, Llama 3.1 8B Instruct is cheaper. More input per output token, as with retrieval or agents that resend long contexts, favours the lower input price; long answers favour the lower output price.
| Input : output tokens | gpt-oss-20b | Llama 3.1 8B Instruct |
|---|---|---|
| 1 : 3 (long answers) | USD 0.072 AkashML | USD 0.035 lower DeepInfra |
| 1 : 1 | USD 0.054 AkashML | USD 0.03 lower DeepInfra |
| 3 : 1 (the site’s blend) | USD 0.036 AkashML | USD 0.025 lower DeepInfra |
| 10 : 1 (retrieval, long prompts) | USD 0.02455 AkashML | USD 0.02182 lower DeepInfra |
| 30 : 1 (agents, big contexts) | USD 0.02032 lower AkashML | USD 0.02065 DeepInfra |
USD per 1M tokens at the cheapest standard provider for each mix, before cache discounts. Thinking tokens bill as output.
What do gpt-oss-20b and Llama 3.1 8B Instruct cost per character in other languages?
Cost to send 1M characters at each model's cheapest non-promotional standard input price, with the gpt-oss (o200k_harmony) tokenizer for gpt-oss-20b and the Llama 3.3 70B tokenizer for Llama 3.1 8B Instruct:
| Content | gpt-oss-20b | Llama 3.1 8B Instruct |
|---|---|---|
| English | USD 0.0064 CoreWeave Forge | USD 0.0043 lower DeepInfra |
| Arabic MSA | USD 0.0096 CoreWeave Forge | USD 0.0077 lower DeepInfra |
| Egyptian Arabic | USD 0.0100 CoreWeave Forge | USD 0.0079 lower DeepInfra |
| Hindi | USD 0.0097 lower CoreWeave Forge | USD 0.010 DeepInfra |
| Chinese | USD 0.023 CoreWeave Forge | USD 0.016 lower DeepInfra |
| Spanish | USD 0.0068 CoreWeave Forge | USD 0.0053 lower DeepInfra |
| Code | USD 0.0076 CoreWeave Forge | USD 0.0051 lower DeepInfra |
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
- GPT OSS 20B API pricing: all 10 providers and other tiers.
- Llama 3.1 8B Instruct API pricing: all 6 providers and other tiers.
- Cost calculator with gpt-oss-20b and Llama 3.1 8B Instruct preloaded
- gpt-oss token counter
- Llama 3 token counter
- Prompt caching calculator
More comparisons: gpt-oss-120b vs gpt-oss-20b · Gemma 4 26B A4B vs gpt-oss-20b · gpt-oss-20b vs Qwen3.6 35B A3B · Llama 3.1 8B Instruct vs Llama 3 8B · Llama 3.1 70B vs Llama 3.1 8B Instruct · every gpt-oss-20b comparison · every Llama 3.1 8B Instruct comparison · all 662 price comparisons
Questions
Is gpt-oss-20b cheaper than Llama 3.1 8B Instruct?
No. On a blend of 3 input tokens to 1 output token, Llama 3.1 8B Instruct costs USD 0.025 per 1M tokens and gpt-oss-20b USD 0.036, at each model's cheapest standard price. For input alone, gpt-oss-20b is cheaper (USD 0.018 against USD 0.02 per 1M); for output alone, Llama 3.1 8B Instruct is (USD 0.04 against USD 0.09). This compares price only, not quality.
What do gpt-oss-20b and Llama 3.1 8B Instruct cost per 1M tokens?
gpt-oss-20b: from USD 0.018 input and USD 0.09 output at AkashML (promotional); median of 10 standard prices USD 0.04 / USD 0.19. Llama 3.1 8B Instruct: from USD 0.02 input and USD 0.04 output at DeepInfra; median of 6 standard prices USD 0.21 / USD 0.22. Each price is read from the provider's own pricing page; the source is linked in the table.
Which providers sell both gpt-oss-20b and Llama 3.1 8B Instruct?
5 providers have a standard price for both: DeepInfra (USD 0.0575 and USD 0.025 at 3:1), CoreWeave Forge (USD 0.055 and USD 0.22 at 3:1), Featherless (USD 0.0675 and USD 0.23 at 3:1), Amazon Bedrock (USD 0.1275 and USD 0.22 at 3:1), Cloudflare Workers AI (USD 0.225 and USD 0.4182 at 3:1). The first is the lowest combined price. Buying each model from its own cheapest provider costs USD 0.036 and USD 0.025.
Which is cheaper for a coding agent or a chatbot, gpt-oss-20b or Llama 3.1 8B Instruct?
For a coding agent (1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning): USD 26.73 a month on gpt-oss-20b (AkashML) against USD 39.60 on Llama 3.1 8B Instruct (DeepInfra), a difference of USD 12.87. For a support chatbot (2,000 requests a day, 1,500 input tokens (60% cached), 300 output): USD 2.75 a month on gpt-oss-20b (AkashML) against USD 2.52 on Llama 3.1 8B Instruct (DeepInfra), a difference of USD 0.23. 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 gpt-oss-20b and Llama 3.1 8B Instruct change?
The cheaper model changes with the mix: below 25 input tokens per output token, Llama 3.1 8B Instruct is cheaper; above 25, gpt-oss-20b is cheaper. At the site's 3:1 blend, Llama 3.1 8B Instruct is cheaper. More input per output token, as with retrieval or agents that resend long contexts, favours the lower input price; long answers favour the lower output price.
Do gpt-oss-20b and Llama 3.1 8B Instruct have cached input pricing?
gpt-oss-20b from USD 0.007 per 1M cached input tokens (Google Vertex AI; 4 of 10 standard prices list a discount). Llama 3.1 8B Instruct from USD 0.028 per 1M cached input tokens (Featherless; 1 of 6 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, gpt-oss-20b or Llama 3.1 8B Instruct?
Both go up to 131,072 tokens on the providers that state it.
Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity
Sources
- AkashML: pricing (checked 2026-11-15)
- Google Vertex AI: pricing (checked 2026-11-14)
- CoreWeave Forge: pricing (checked 2026-11-15)
- DeepInfra: pricing (checked 2026-11-14)
- Featherless: pricing (checked 2026-11-15)
- DeepInfra: pricing (checked 2026-11-14)
- Featherless: pricing (checked 2026-11-15)
- Amazon Bedrock: pricing (checked 2026-11-14)
- Cloudflare Workers AI: pricing (checked 2026-11-14)