Phi-3-medium-128k-instruct: benchmark scores, price and ranking

Phi-3-medium-128k-instruct from Microsoft ranks #653 of 743 LLMs on the LLMs Tiger aggregate benchmark score (44.0), based on 2 benchmark results captured up to 2026-10-13.

Organisation: Microsoft
Type: Open-weight / local model
Open weights: Yes
Released: 2024-04-23 (source: epoch)
Cheapest API price: $0.170 in / $0.680 out per million tokens (via azure_ai, observed 2026-10-13T20:47Z)
Aggregate rank: #653 of 743 (score 44.0)

Across the 2 benchmarks where at least ten models are scored, Phi-3-medium-128k-instruct is in the top 10% on 0 and in the top half on 0.

Its cheapest listed input price, $0.170 per million tokens, is lower than that of 71% of the 427 models with a listed price.

Phi-3-medium-128k-instruct benchmark results

Phi-3-medium-128k-instruct benchmark scores, ranked against all models measured on each benchmark
Benchmark	Score	Rank	Source	Captured
MMLU-Pro	51.9%	#75 of 136	HuggingFace MMLU-Pro leaderboard	2026-10-13
GPQA Diamond	27.6%	#256 of 275	Epoch AI Benchmarking Hub	2026-10-13

About Phi-3-medium-128k-instruct

Phi-3-medium-128k-instruct is an open-weight model that can run on your own hardware from Microsoft. It was released on 2024-04-23, according to Epoch AI. The cheapest API listing we track charges $0.170 per million input tokens and $0.680 per million output tokens, through azure_ai. We hold 2 benchmark results for it across 1 area, and none is in the top tenth of its board.

Where Phi-3-medium-128k-instruct stands

On our aggregate score, Phi-3-medium-128k-instruct is #653 of 743 ranked models, in the lower half. Its score is 44.0; the scale is centred on 50, so it sits 6.0 points below the typical model. That beats 12% of the ranked field. Among the 3 ranked Microsoft models it is #3. The next one up is WizardLM-2-8x22B at 45.3.

Phi-3-medium-128k-instruct, benchmark by benchmark

Phi-3-medium-128k-instruct lands below the middle on MMLU-Pro: 51.9%, #75 of 136, with 54% of the field scoring higher. Darwin-180B-RSI leads at 88.1%; the closest result above is Yi-1.5-34B-Chat at 52.3%. It covers 57-subject knowledge + reasoning. The board median is 55.3%, so Phi-3-medium-128k-instruct is 3.3 points below it. Source: HuggingFace MMLU-Pro leaderboard, captured 2026-10-13.

GPQA Diamond is a weak spot. Phi-3-medium-128k-instruct scores 27.6%, #256 of 275, behind 93% of the models measured. The leader, GPT-6 Astra (max), reaches 95.8%. It covers expert-level science questions. The board median is 71.7%, so Phi-3-medium-128k-instruct is 44.1 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Strengths and weaknesses

Reasoning, maths and science: weak, ahead of 26% of the field on average across 2 benchmarks. Its best board here is MMLU-Pro (#75 of 136) and its weakest is GPQA Diamond (#256 of 275).

Price and value

The median input price across the 427 priced models is $0.500 per million tokens, so Phi-3-medium-128k-instruct costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen2.5-Coder-3B-Instruct ($0.010, 46.5). Similar scores cost more on DeepSeek-V3 ($0.200) and grok-code-fast-1 ($0.200).

Release history

The Microsoft model released before it was WizardLM-2-8x22B on 2024-04-15, which scores 45.3. The next one was Phi-3-mini-4k-instruct on 2024-04-23, at 46.2.

Reading these numbers

The 2 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (1), HuggingFace MMLU-Pro leaderboard (1). No official leaderboard supplied any of them, so treat them as third-party measurements. With 2 results the aggregate rank can move a lot when one new score arrives.

Models ranked nearby

#650 gemini-1.5-pro-001 (44.0)
#651 Mellum2-12B-A2.5B-Base (44.0)
#652 Mellum2-12B-A2.5B-Base-Pretrain (44.0)
#654 Yi 6B (44.0)
#655 Mistral Small 3.2 (44.0)
#656 Hermes-2-Theta-Llama-3-70B (44.0)