Spiral gives your agent every tool with one account. Browse the tools →

Codex · LLMReleased 2024-07-23

How to Use Llama 3.1 70B Instruct in Codex

Llama 3.1 70B Instruct handles document summaries, multilingual support, and coding tasks. Meta recommends it over Llama 3.1 8B Instruct when conversations also need tool calls.

Prompts to try with Llama 3.1 70B Instruct in Codex

Copy an example into Codex, or ask in your own words.

Use Llama 3.1 70B Instruct to inspect the failing test and the implementation it exercises. Propose the smallest fix, explain the cause, and identify the test result that would confirm it. Keep the explanation under 400 words.

Llama 3.1 70B InstructCodex
Try asking
  • Use Llama 3.1 70B Instruct to review the current diff for behavior changes and missing edge cases. Cite file paths and line numbers for each finding. Separate confirmed problems from questions.

  • Use Llama 3.1 70B Instruct to turn this project's recent commit descriptions into release notes. List user-visible changes first, then maintenance work. Flag changes whose purpose cannot be established from the code.

  • Use Llama 3.1 70B Instruct to draft Spanish and German versions of the English help text in this project. Preserve product names, interpolation tokens, and code examples. List ambiguous phrases for review.

  • Use Llama 3.1 70B Instruct to generate 12 fictional support conversations from this project's help documentation. Include four cases where the documentation cannot answer the question. Follow the existing evaluation format and mark the expected supporting passage.

How to set up Llama 3.1 70B Instruct in Codex

Pick the way that suits you. Each takes about two minutes.

Use Spiral's MCP server

  1. Add the hosted server
    codex mcp add spiral --url https://api.spiral.ai/mcp
  2. Sign in

    Codex opens the browser: sign in to Spiral and click Authorize. No API key. codex mcp list then shows Spiral.

    codex mcp login spiral

Use the Spiral skill

  1. Send Codex this message

    Your agent reads Spiral's skill and sets itself up.

    set up https://spiral.ai/SKILL.md
  2. Sign in or paste a key

    Approve Spiral in the browser, or give the agent a key from the console.

Use the Spiral CLI

  1. Install and sign in

    Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

    npx -y spiral-cli@latest init

Call the Spiral API

  1. Create an API key

    In the console, under API keys.

  2. Run Llama 3.1 70B Instruct

    One request; the result comes back by polling or a webhook.

    curl https://api.spiral.ai/v1/runs \
      -H "Authorization: Bearer $SPIRAL_API_KEY" \
      -H "Content-Type: application/json" \
      -H "Idempotency-Key: $(uuidgen)" \
      -d '{
        "provider": "openrouter",
        "model": "meta-llama/llama-3.1-70b-instruct",
        "input": {
          "messages": [
            {
              "role": "system"
            }
          ]
        }
      }'

Then paste one of the prompts above: Codex finds Llama 3.1 70B Instruct on Spiral, shows the price and runs it.

Llama 3.1 70B Instruct providers and prices

Spiral runs Llama 3.1 70B Instruct through OpenRouter. Open a card for today's price, with every fee included.

OpenRouterLLM
Meta: Llama 3.1 70B Instruct
meta-llama/llama-3.1-70b-instruct

Settings you can ask for

Ask Codex for any of these in your prompt.

SettingAllowed valuesDefaultDescription
Top ka whole number from 0
Top p0 to 1
Max tokens1 to 8192

The most output tokens to generate.

Temperature0 to 2

Every API field of Meta: Llama 3.1 70B Instruct is on its tool page: open its card above.

What can you do with Llama 3.1 70B Instruct in Codex?

Turn project code, error logs, and technical notes into proposed fixes and readable explanations with Llama 3.1 70B Instruct. Include the relevant files and ask for a focused change plus a way to verify it.

Decision briefs from long documents

Turn a lengthy proposal or report into a short brief with decisions, supporting passages, and unresolved questions. Request section references so each conclusion can be checked.

Multilingual customer support drafts

Draft replies in the eight supported languages using supplied product information. Preserve product names and separate documented answers from questions that need a support specialist.

Code review and debugging assistance

Review relevant code and error output to propose a focused fix. Ask for the suspected cause, the proposed change, and a test that would confirm the fix.

Meeting summaries with traceable actions

Convert a text transcript into decisions, action items, and open questions. Require transcript evidence for each assigned owner or deadline.

Answers grounded in reference material

Answer questions using supplied help articles or policy excerpts. Request a supporting passage for every answer and an explicit statement when the material does not cover the question.

Support conversation test data

Generate fictional support conversations for evaluation, including follow-up questions and cases with missing information. Review the examples before using them to assess another system.

Release notes from technical changes

Turn commit descriptions and engineering notes into a readable release summary. Keep user-visible changes separate from internal maintenance and flag claims that need confirmation.

Questions about Llama 3.1 70B Instruct in Codex

Can Codex use Llama 3.1 70B Instruct?

Yes. Add Spiral to Codex as an MCP server, and Codex can find and run Llama 3.1 70B Instruct like any built-in tool.

How much does Llama 3.1 70B Instruct cost in Codex?

Up to $0.003868 per run at the default inputs. $0.936 per 1M input tokens, $0.936 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Codex plan is separate.

Do I need an OpenRouter account to use Llama 3.1 70B Instruct in Codex?

No. Spiral runs Llama 3.1 70B Instruct on its own OpenRouter account. You sign in to Spiral once and never paste a key.

How long does it take to set up Llama 3.1 70B Instruct in Codex?

About two minutes: add the hosted server, then sign in.

Does Codex wait while Llama 3.1 70B Instruct runs?

Llama 3.1 70B Instruct runs in the background. Codex starts the run, checks on it and gives you the text, json when it's ready.

What is Llama 3.1 70B Instruct useful for?

Use it to summarize lengthy text, draft multilingual support replies, and assist with code. Give it the relevant source material and a concrete deliverable, such as a decision list or proposed patch.

How much text can Llama 3.1 70B Instruct consider?

Meta specifies a 128K-token context length. That figure describes context capacity, so do not treat it as a promise of a 128K-token response; set a separate response-length target.

Which languages does Llama 3.1 70B Instruct support?

Meta lists English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai. Specify the output language and any terminology that must stay unchanged.

Can Llama 3.1 70B Instruct interpret images or generate video?

It is a text model. For a task based on a screenshot or recording, provide a text description or transcript.

What should I check in its answers?

Check factual claims against the supplied sources, especially when an answer sounds certain. Ask it to quote supporting passages and mark missing information, then verify those passages yourself.

Get started

Start using Llama 3.1 70B Instruct in Codex

One Spiral account for every tool your agent needs. No subscription; you pay per run.

Paste this into Codex:

Set up Codex

set up https://spiral.ai/SKILL.md
Are you an AI agent? See setup options