All models

Claude 3 Haiku vs Llama 3.1 70B Instruct

Claude 3 Haiku at $0.287 in and $1.44 out and Llama 3.1 70B Instruct at $0.460 in and $0.460 out — per million tokens, from the same wallet.

Claude 3 Haiku

Anthropic

0.26¢

for a client proposal

$0.287 in · $1.44 out / 1M

200K context · released March 2024

Full page →
M

Llama 3.1 70B Instruct

Meta

0.27¢

for a client proposal

$0.460 in · $0.460 out / 1M

131K context · released July 2024

Full page →

Asking all both at once — identical prompt, identical context — costs about 0.53¢ for a typical client proposal.

At a glance

  • Cheapest input: Claude 3 Haiku at $0.287 per million tokens.
  • Cheapest output: Llama 3.1 70B Instruct at $0.460 per million tokens.
  • Largest context: Claude 3 Haiku at 200K tokens — about 290 pages.

Cost per task

Estimated totals at the rates above. Solid bar is input, lighter bar is output.

Quick code fix

2K in · 500 out
Claude 3 Haiku
0.13¢
Llama 3.1 70B Instruct
0.11¢

Draft a client proposal from notes

5K in · 800 out
Claude 3 Haiku
0.26¢
Llama 3.1 70B Instruct
0.27¢

Context-heavy session (Alyph workspace)

155K in · 4K out
Claude 3 Haiku
$0.05
Llama 3.1 70B Instruct
needs 155K ctx

Analyze a 50-page legal contract

35K in · 1K out
Claude 3 Haiku
1.1¢
Llama 3.1 70B Instruct
1.7¢

Large architecture refactor

250K in · 3K out
Claude 3 Haiku
needs 250K ctx
Llama 3.1 70B Instruct
needs 250K ctx

Solid part of each bar is input tokens, the lighter part is output. Models that don’t fit a task are marked instead of priced.

Cost vs Token Amount

Drag the slider to adjust the split between input and output workload.

80% Input20% Output
free$0.26$0.520 tokens250k500k750k1000k
Claude 3 Haiku
Llama 3.1 70B Instruct

Adjust the ratio slider to change how output tokens influence the final price for equivalent workloads.

Pricing

Claude 3 Haiku
Llama 3.1 70B Instruct
Input / 1M tokens
$0.287
$0.460
Output / 1M tokens
$1.44
$0.460
Cached input / 1M
$0.035
—
Web search
1.1¢ / search
—

Specs

Claude 3 Haiku
Llama 3.1 70B Instruct
Provider
Anthropic
Meta
Released
March 2024
July 2024
Context window
200K
131K
Max output
4K
16K
Input
Text and Images
Text
Output
Text
Text
Reasoning
Answers directly
Answers directly
Knowledge cutoff
2023-08-31
2023-12-31

Try it with real numbers

2K tokens in · 500 tokens out.

Claude 3 Haiku

Anthropic

0.13¢

for this task

44% input · 56% output

Full pricing →

Llama 3.1 70B Instruct

Meta

0.11¢

for this task

80% input · 20% output

Full pricing →

All 2, same prompt and context: 0.24¢. That is the entire cost of the comparison.

Your $5 welcome credit covers about 2,046 of these.

Which should you pick?

For long documents, Claude 3 Haiku has the largest context window here — 200K tokens.

On price, Claude 3 Haiku is the cheapest of the two on a typical task, at about 0.26¢.

For images or files, Claude 3 Haiku can read them directly; Llama 3.1 70B Instruct is text-only.

Or don’t pick. Send the identical prompt to all both, read the answers side by side, and keep the winner. How to run a fair bake-off →

Claude 3 Haiku pricing →Llama 3.1 70B Instruct pricing →Add Lyria 3 Clip PreviewAdd Grok 4.20

Questions, answered

Which is cheaper, Claude 3 Haiku or Llama 3.1 70B Instruct?

On a typical client proposal (5K tokens in, 800 out): Claude 3 Haiku at 0.26¢; Llama 3.1 70B Instruct at 0.27¢. Long prompts can change the order when long-context rates apply.

Which has the largest context window?

Claude 3 Haiku — 200K tokens, about 290 pages.

Can I run Claude 3 Haiku and Llama 3.1 70B Instruct side by side on Alyph?

Yes — that is what Alyph is built for: the identical prompt with identical context to all 2, answers rendered side by side, billed from one wallet. This exact combination costs about 0.53¢ for a typical task.

Ask both at once

Identical prompt, identical context, answers side by side — about 0.53¢ for a typical client proposal. Free to start, with $5 of credit.

Sign up free Try the demo