claude-code + claude-haiku-4-5
claude-code harness · Claude Max 5x
overall pending Two distinct suite families are required for an overall grade.
Per-suite results
Run on claude-code 2.1.215 (Claude Code), model claude-haiku-4-5, Jul 19, 2026.
One run of this suite consumes 2.0% of Claude Max 5x - about 50.0 runs per 5h. Measured per suite run (reported-delta); an absolute cost on this plan, not comparable with another agent's plan. Measured in a declared quiet window on a shared subscription: operator Claude Max 5x account on eb-runner, shared with the operator IDE panes; quiet-window until a dedicated benchmark login exists.
Trajectory
Every scored run of this agent, oldest first. Grades compare only within a suite version, so each version gets its own line; while a suite is being retired both versions run, and the superseded one is dashed. The filled point is the run the scorecard shows today. All agents over time →
Jul 19, 2026: A on v1
Compare
- claude-code + claude-haiku-4-5 vs Claude Sonnet 5 (CLI)
- claude-code + claude-haiku-4-5 vs claude-code + claude-fable-5
- claude-code + claude-haiku-4-5 vs claude-code + claude-opus-4-8
- claude-code + claude-haiku-4-5 vs claude-code + claude-sonnet-5
- claude-code + claude-haiku-4-5 vs codex-cli + gpt-5.6-luna
- claude-code + claude-haiku-4-5 vs codex-cli + gpt-5.6-sol
- claude-code + claude-haiku-4-5 vs codex-cli + gpt-5.6-terra
- claude-code + claude-haiku-4-5 vs cursor-cli + composer-2.5
- claude-code + claude-haiku-4-5 vs grok-cli + grok-4.5
- claude-code + claude-haiku-4-5 vs kimi-cli + kimi-k3
Transcript excerpts
Outputs are shown only when they can be separated safely from the private suite. Scores and run metadata remain visible when an output is withheld.
| Task | # | Turns | Latency | Output |
|---|---|---|---|---|
| a1b2c3 | 1 | - | 4442 ms | withheldPrivate-suite content |
| a3f7c8 | 1 | - | 50395 ms | withheldPrivate-suite content |
| a6f1d9 | 1 | - | 10058 ms | output only```json
{
"claims": [
{
"claim": "The battery lasts 12 hours on a single charge.",
"metric": 12,
|
| a9d4f2 | 1 | - | 7000 ms | withheldPrivate-suite content |
| b1d7c3 | 1 | - | 17712 ms | withheldPrivate-suite content |
| b2c9e1 | 1 | - | 12719 ms | withheldPrivate-suite content |
| b8e3d7 | 1 | - | 9210 ms | withheldPrivate-suite content |
| c3f8a4 | 1 | - | 8372 ms | withheldPrivate-suite content |
| c4a9f1 | 1 | - | 12016 ms | withheldPrivate-suite content |
| c8e2a5 | 1 | - | 6804 ms | withheldPrivate-suite content |
| d1e8b6 | 1 | - | 16104 ms | withheldPrivate-suite content |
| d4e5f6 | 1 | - | 4123 ms | withheldPrivate-suite content |
| d6b2e9 | 1 | - | 7620 ms | withheldPrivate-suite content |
| d9c4b2 | 1 | - | 5296 ms | output only```json
{
"records": [
{"id": 1, "name": "Marlin Crate", "amount": 10.5},
{"id": 2, "name": "Harbor Bolt", "am |
| e2b7c5 | 1 | - | 12236 ms | withheldPrivate-suite content |
| e4b9d1 | 1 | - | 8495 ms | withheldPrivate-suite content |
| e7b1c9 | 1 | - | 11984 ms | withheldPrivate-suite content |
| f0a7d2 | 1 | - | 20365 ms | withheldPrivate-suite content |
| f3a8e6 | 1 | - | 10908 ms | withheldPrivate-suite content |
| f5c2a8 | 1 | - | 6874 ms | withheldPrivate-suite content |