fix(pricing): Claude Haiku 4.5 actual Anthropic pricing ($0.25 → $1.00/M) by engkimo · Pull Request #37 · engkimo/open-morphic

engkimo · 2026-05-19T00:51:36Z

Summary

Both Morphic cost tables had Haiku 4.5 priced at Haiku 3.5-era levels ($0.25 input / $1.25 output per 1M tokens). Anthropic's actual Haiku 4.5 public pricing is $1.00 input / $5.00 output per 1M tokens.

This made every Haiku-tier cost projection (including yesterday's #36 planner cost simulation) understate real spend by ~4× on input and on output.

Changes

application/use_cases/cost_estimator.py — MODEL_COST_TABLE input: $0.25 → $1.00/M
domain/services/engine_cost_calculator.py — _MODEL_PRICING: ($0.25, $1.25) → ($1.00, $5.00)/M
Refreshed "Updated YYYY-MM" comment

Impact on cost simulation

Re-running benchmarks/planner_cost_simulation.py with corrected pricing:

Sonnet → Haiku savings drop from −66.7% → −67% (still very strong; Haiku output is now $5 vs Sonnet $15 → 3× output saving instead of 3× — Haiku stays decisively cheaper)
Padding conclusion unchanged (still net-negative for short-prompt workloads)
Recommendation unchanged: switch planner default to Haiku 4.5 when Claude tier is opted into

Test plan

uv run --extra dev pytest tests/unit/ -q — 3,235 passed, 0 regressions
uv run --extra dev ruff check . — clean
test_haiku_cheaper_than_sonnet invariant still holds ($1 < $3)

🤖 Generated with Claude Code

Summary by CodeRabbit

Chores
- Updated pricing information for the Claude Haiku model. Input and output token costs have been revised to reflect current market rates, which will affect cost calculations and budget estimates for operations using this model.

Both cost tables had Haiku 4.5 at Haiku 3.5-era prices ($0.25/$1.25 per 1M tokens). Actual Anthropic 4.5 pricing is $1.00/$5.00. This made every Haiku-tier cost projection (incl. the planner cost simulation) understate real spend by 4×. - application/use_cases/cost_estimator.py: input $0.25 → $1.00 - domain/services/engine_cost_calculator.py: ($0.25, $1.25) → ($1.00, $5.00) - All 3,235 unit tests still pass (haiku < sonnet invariant intact: $1 < $3) Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

coderabbitai · 2026-05-19T00:51:46Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: a7b3d983-d61e-4f67-ae36-7caaa3833ce3

📥 Commits

Reviewing files that changed from the base of the PR and between f86706a and fd423c1.

📒 Files selected for processing (2)

application/use_cases/cost_estimator.py
domain/services/engine_cost_calculator.py

📝 Walkthrough

Walkthrough

This PR updates token pricing for the claude-haiku-4-5-20251001 Anthropic model across two cost-calculation modules. The cost estimator's input token rate increases from 0.25 to 1.00 per 1M tokens, while the engine cost calculator applies a more comprehensive update to both input and output rates (1.00 and 5.00 respectively), and refreshes the pricing verification timestamp to May 2026.

Changes

Model Pricing Update

Layer / File(s)	Summary
claude-haiku-4-5-20251001 pricing update `application/use_cases/cost_estimator.py`, `domain/services/engine_cost_calculator.py`	Input token cost per 1M tokens raised from 0.25 to 1.00 in the cost estimator. Engine cost calculator updated with new per-1M-token rates of (1.00 input, 5.00 output) replacing (0.25 input, 1.25 output), and pricing verification timestamp updated from 2026-03 to 2026-05-19.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Poem

🐰 Prices rise like spring's fresh carrots,
From 0.25 to 1.00, the tokens now soar,
Haiku speaks at higher rates,
Two tables aligned in harmony's dance,
May's pricing blooms anew! 🥕✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately reflects the main change: updating Claude Haiku 4.5 pricing from $0.25 to $1.00 per million input tokens to match Anthropic's actual pricing.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

Create stacked PR
Commit on current branch

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch fix/haiku-4-5-pricing

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

- Promote Unreleased CHANGELOG entry to v0.6.3 (2026-05-22), rolling up 7 PRs since v0.6.2: TD-195 router (#40), plan-evaluator fix (#39), Haiku A/B (#38), Haiku pricing fix (#37), cost-sim harness (#36), TD-189 step 5 CLI (#35). - Bump pyproject + main.py + version consistency test + uv.lock. - Refresh CLAUDE.md version banner to 0.6.3 / 2026-05-22.

engkimo merged commit 8ccb3f3 into main May 19, 2026
6 checks passed

engkimo deleted the fix/haiku-4-5-pricing branch May 19, 2026 00:54

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix(pricing): Claude Haiku 4.5 actual Anthropic pricing ($0.25 → $1.00/M)#37

fix(pricing): Claude Haiku 4.5 actual Anthropic pricing ($0.25 → $1.00/M)#37
engkimo merged 1 commit into
mainfrom
fix/haiku-4-5-pricing

engkimo commented May 19, 2026 •

edited by coderabbitai Bot

Loading

Uh oh!

coderabbitai Bot commented May 19, 2026 •

edited

Loading

Walkthrough

Changes

Estimated code review effort

Poem

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

engkimo commented May 19, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Changes

Impact on cost simulation

Test plan

Summary by CodeRabbit

Uh oh!

coderabbitai Bot commented May 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Poem

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

engkimo commented May 19, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented May 19, 2026 •

edited

Loading