Head to head
Phi 4vsLlama 3.1 8B Instruct
For tasks, about a month.
Verdict · measured basis
Llama 3.1 8B Instruct costs $0.0000 per solved task against
Phi 4 at $0.0000 — ▼ 1.0x cheaper.
Over 200 tasks that is $0.00 a month.
On output token price alone Llama 3.1 8B Instruct looks 1.8x cheaper. The ranking holds, but the margin changes.
Verified pricing, pass rate and cost per solved task for Phi 4 and Llama 3.1 8B Instruct| Figure | Phi 4 | Llama 3.1 8B Instruct |
|---|
| Input $/Mtok | $0.07 | $0.05 |
|---|
| Output $/Mtok | $0.14 | $0.08 |
|---|
| Cached input $/Mtok | $0.07 | $0.025 |
|---|
| Pass rate | 67% | 42% |
|---|
| Cost per solved task | $0.0000measured | $0.0000measured |
|---|
| Cost per month | $0.01 | $0.01 |
|---|
Cost per month against tasks per month · drag the marker
Solid measured · dashed modeled · dotted stale. The gap is the verdict.
Solvency Bench (solvency.dev) — first-party measurement · verified 2026-08-26 · prices from each provider's pricing page; measured rows ignore the tier · Methodology · All models at these settings