Evidence in this edition2 source links

none reviewed yet · 2 benchmark observations · latest capture Nov 15, 2026

Inspect the evidence
Task fitUnscored

No editorial score yet — missing data is not a low score

How scoring works
Cost at your workload$85.20

Verified rates (Nov 15, 2026) · 100 runs / month at your token settings

Adjust workload

At a glance

9 facts
Provider
Meta
Type
API model
Access
OpenRouter API
Context
128,000 tokens
Maximum output
16,384 tokens
License / weights
Open weights
Input modalities
text
Output modalities
text
Pricing notes
$0.1 input / $0.32 output per 1M tokens on OpenRouter.
From the facts on this page

Good to know

In its favour

  • Listed with tool calling and structured output.
  • Open weights: it can run on your own hardware (licence: n.a.).

Check before you rely on it

Nothing flagged by the recorded facts.

Worked out from the catalogue’s own facts and dates by the same rule for every model. Not a review, and no one pays to change it.

At its verified price

Example bills

Same token sizes as the cheapest-by-task pages, so you can see where this model ranks. Jobs that do not fit its context window are left out.

Price ledger

Price history

2 readings
Output price per 1M tokens (USD)
Sep 11, 2026: $0.10 input / $0.32 output per 1M tokens (OpenRouter)Nov 15, 2026: $0.71 input / $0.71 output per 1M tokens (OpenRouter)$0.32$0.71Sep 11, 2026Nov 15, 2026
ReadPrice per 1M (in / out)Source
Nov 15, 2026$0.71 / $0.71OpenRouter
Sep 11, 2026$0.10 / $0.32OpenRouter

Newest first, in the order the catalogue recorded them. A new row appears only when the price, its source or its status changed. All price changes

The practical trade-offs

Documented strengths

  • Current model catalogue entry with a public API identifier and pricing metadata.

Limitations to check

  • No normalized frontend-quality score has been assigned.
  • Popularity and recency are not treated as quality evidence.
Follow the source

Evidence & conditions

2 source links
2Source links
NoneReviewed records
2Benchmark runs
Nov 15, 2026Latest capture

Benchmark observations

HumanEval Inspect conditions
name
HumanEval
value
88.4
unit
%
configuration
Llama 3.3 70B Instruct; zero-shot; pass@1.
attribution
provider-reported
source Url
https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
checked At
2026-09-08T19:31:52Z
MBPP EvalPlus (base) Inspect conditions
name
MBPP EvalPlus (base)
value
87.6
unit
%
configuration
Llama 3.3 70B Instruct; zero-shot; pass@1; base evaluation subset.
attribution
provider-reported
source Url
https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
checked At
2026-09-08T19:31:52Z

Source reports, measured results, and editorial assessments are separate evidence types. Source dates do not imply a new benchmark run. Capture dates are shown when recorded; otherwise we say the capture date was not recorded. Recorded dates are compared with this catalogue snapshot (Nov 15, 2026), not today’s date.

How we rate

How this page is assessed

Quality claims follow how we rate (method AIS-V2).

Money never buys rank — commercial relationships cannot change rank, eligibility, evidence quality, or publication timing.

Submit or claim a listing is free and does not skip the editorial queue.

Private · this device only

Your research note

Saved on this device. Included only in private backups.