Skip to content
AI Atlas
Model

Qwen2.5-Coder

Qwen Teamfamily · Qwen2.5-Coderqwen.ai

Open in Graph
data quality61

Updated 5 h ago · first seen 12 Sept 2026

model_01M29X37HMXPWD8GV80H5V8YXK

Overview

Identity

Identity block not returned by the API for this entity.

Openness

Openness not classified yet — no sourced evidence to place this model in the ontology.

Key facts

Release date

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Status

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Version

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Official page
qwen.ai

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Paper
arxiv.org/abs/2409.12186

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Architecture

Architecture

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Parameters

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Active parameters

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Mixture of experts

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Capabilities

Modalities

Modalities
text
Input
text
Output
text

Capabilities

  • Tool calling

    Unavailable

  • Structured output

    Unavailable

  • Reasoning

    Yes

    Qwen — official blog · T2

  • Vision

    No

    Qwen — official blog · T2

  • Audio

    No

    Qwen — official blog · T2

  • Fine-tuning available

    Yes

    Qwen — official blog · T2

Context window

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Languages

Source:Qwen — official blogT2observed 13 h agomediumLLM-extracted

Hardware fit37

Estimated

36 of 37 device × quantization combinations fit.

Run locally: your machine →
Estimated hardware fit
HardwareQuantizationDevice memoryEst. memoryFits
Apple M3 Ultra4bit—19.2 GB est.Yes
Mac Studio (Apple M5 Ultra)4bit—19.2 GB est.Yes
Apple M2 Ultra4bit—19.2 GB est.Yes
Apple M1 Ultra4bit—19.2 GB est.Yes
Apple M3 Max4bit—19.2 GB est.Yes
Apple M4 Max4bit—19.2 GB est.Yes
Mac Studio (Apple M5 Max)4bit—19.2 GB est.Yes
MacBook Pro (Apple M5 Max)4bit—19.2 GB est.Yes
Apple M2 Max4bit—19.2 GB est.Yes
Apple M1 Max4bit—19.2 GB est.Yes
Apple M4 Pro4bit—19.2 GB est.Yes
Mac mini (Apple M5 Pro)4bit—19.2 GB est.Yes
MacBook Pro (Apple M5 Pro)4bit—19.2 GB est.Yes
Apple M3 Pro4bit—19.2 GB est.Yes
Apple M1 Pro4bit—19.2 GB est.Yes
Apple M2 Pro4bit—19.2 GB est.Yes
Apple M44bit—19.2 GB est.Yes
iMac (Apple M4)4bit—19.2 GB est.Yes
Mac mini (Apple M6)4bit—19.2 GB est.Yes
MacBook Air (Apple M5)4bit—19.2 GB est.Yes
MacBook Pro (Apple M5)4bit—19.2 GB est.Yes
Apple M24bit—19.2 GB est.Yes
Apple M34bit—19.2 GB est.Yes
NVIDIA GeForce RTX 30904bit24 GB19.2 GB est.Yes
NVIDIA GeForce RTX 40904bit24 GB19.2 GB est.Yes
NVIDIA GeForce RTX 50904bit32 GB19.2 GB est.Yes
NVIDIA A100 80GB4bit80 GB19.2 GB est.Yes
NVIDIA H100 SXM4bit80 GB19.2 GB est.Yes
NVIDIA H100 NVL4bit94 GB19.2 GB est.Yes
NVIDIA DGX Spark4bit128 GB19.2 GB est.Yes
NVIDIA H2004bit141 GB19.2 GB est.Yes
NVIDIA H200 NVL4bit141 GB19.2 GB est.Yes
NVIDIA B2004bit180 GB19.2 GB est.Yes
AMD Instinct MI300X4bit192 GB19.2 GB est.Yes
AMD Instinct MI325X4bit256 GB19.2 GB est.Yes
NVIDIA DGX B2004bit1,440 GB19.2 GB est.Yes
Apple M14bit—19.2 GB est.No
Assumptions (6)
  • Estimated, not measured: weights = parameters × bytes/param × 1.15 runtime overhead.
  • bytes/param: 4bit = 0.5, 8bit = 1.0, fp16 = 2.0 (uniform quantization, no per-layer exceptions).
  • KV cache approximated at 0.5 GB per 8 192 tokens of context, independent of architecture (GQA/MLA models need less).
  • A model 'fits' when the estimate is at most the device memory minus 2 GB reserved for the OS and framework.
  • Mixture-of-experts models are estimated on total parameters (all experts must be resident); active parameters are ignored.
  • Device memory uses the largest configuration when several are listed (e.g. Apple silicon tiers).

Lineage

Open in Graph →
Explicit derived_from / fine_tuned_from / distilled_from relations stated by sources; artifacts collapsed by kind.
ANCESTORS 3Qwen2.572.7BQwen2.5 — 72.7BQwen2.5-Coder-BaseQwen2.5-Coder-BaseQwen2Qwen2Qwen2.5-Coder32.5B params · this modelQwen2.5-Coder — 32.5B params · this model

Timeline7

Full timeline →
  • New modelModelQwen2.5-CoderQwen Team

    New model: Qwen2.5-Coder (Qwen Team)

    qwen
    qwenlm.github.io
  • Property changedModelQwen2.5-CoderQwen Team

    Qwen2.5-Coder: languages changed from ["92 programming languages"] to ["Chinese", "English"]

    Languages92 programming languages→Chinese, Englishqwen
    qwenlm.github.io
  • Property changedModelQwen2.5-CoderQwen Team

    Qwen2.5-Coder: architecture changed from transformer to Transformer

    Architecturetransformer→Transformerqwen
    qwenlm.github.io
  • License changedModelQwen2.5-CoderQwen Team

    Qwen2.5-Coder: license changed from Apache 2.0 to Apache 2.0 (for 0.5B, 1.5B, 7B, 14B, 32B); Qwen-Research …

    LicenseApache 2.0→Apache 2.0 (for 0.5B, 1.5B, 7B, 14B, 32B); Qwen-Research (for 3B)qwen
    qwenlm.github.io
  • Status changedModelQwen2.5-CoderQwen Team

    Qwen2.5-Coder: status changed from active to open-source

    Statusactive→open-sourceqwen
    qwenlm.github.io
  • Release date changedModelQwen2.5-CoderQwen Team

    Qwen2.5-Coder: release date changed from 2024-09-19 to 2024-11-12

    Release date19 Sept 2024→12 Nov 2024qwen
    qwenlm.github.io
  • Property changedModelQwen2.5-CoderQwen Team

    Qwen2.5-Coder: family changed from Qwen2.5 to Qwen2.5-Coder

    FamilyQwen2.5→Qwen2.5-Coderqwen
    qwenlm.github.io

Change history28

Temporal, append-only claims: a new observation closes the previous claim instead of overwriting it. Rewind the record with the as-of picker.

Viewing AI Atlas as of 1 Jan 2025 — attributes exactly as the atlas knew them on that day; later corrections are not shown.

Back to today →

Qwen2.5-Coder was not yet in AI Atlas on 1 Jan 2025

First seen 12 Sept 2026. Nothing is inferred backwards: no attribute is shown for dates before the first observation.
28 claims · 22 properties

Familyfamily2

Claim history for Family
ValueValid from → toStatusSourceConfidenceExtractor
Qwen2.5 → supersededQwen — official blogT2mediumllm
Qwen2.5-Coder → currentcurrentQwen — official blogT2mediumllm

Versionversion1

Claim history for Version
ValueValid from → toStatusSourceConfidenceExtractor
2.5 → currentcurrentQwen — official blogT2mediumllm

Release daterelease_date2

Claim history for Release date
ValueValid from → toStatusSourceConfidenceExtractor
19 Sept 2024 → supersededQwen — official blogT2mediumllm
12 Nov 2024 → currentcurrentQwen — official blogT2mediumllm

Statusstatus2

Claim history for Status
ValueValid from → toStatusSourceConfidenceExtractor
active → supersededQwen — official blogT2mediumllm
open-source → currentcurrentQwen — official blogT2mediumllm

Opennessopenness1

Claim history for Openness
ValueValid from → toStatusSourceConfidenceExtractor
open-source → currentcurrentQwen — official blogT2mediumllm

Licenselicense2

Claim history for License
ValueValid from → toStatusSourceConfidenceExtractor
Apache 2.0 → supersededQwen — official blogT2mediumllm
Apache 2.0 (for 0.5B, 1.5B, 7B, 14B, 32B); Qwen-Research (for 3B) → currentcurrentQwen — official blogT2mediumllm

Architecturearchitecture2

Claim history for Architecture
ValueValid from → toStatusSourceConfidenceExtractor
transformer → supersededQwen — official blogT2mediumllm
Transformer → currentcurrentQwen — official blogT2mediumllm

Parametersparameter_count1

Claim history for Parameters
ValueValid from → toStatusSourceConfidenceExtractor
32.5B → currentcurrentQwen — official blogT2mediumllm

Active parametersactive_parameter_count1

Claim history for Active parameters
ValueValid from → toStatusSourceConfidenceExtractor
31B → currentcurrentQwen — official blogT2mediumllm

Mixture of expertsis_moe1

Claim history for Mixture of experts
ValueValid from → toStatusSourceConfidenceExtractor
No → currentcurrentQwen — official blogT2mediumllm

Context windowcontext_length1

Claim history for Context window
ValueValid from → toStatusSourceConfidenceExtractor
128K tokens → currentcurrentQwen — official blogT2mediumllm

Modalitiesmodalities1

Claim history for Modalities
ValueValid from → toStatusSourceConfidenceExtractor
text → currentcurrentQwen — official blogT2mediumllm

Input modalitiesmodalities_input1

Claim history for Input modalities
ValueValid from → toStatusSourceConfidenceExtractor
text → currentcurrentQwen — official blogT2mediumllm

Output modalitiesmodalities_output1

Claim history for Output modalities
ValueValid from → toStatusSourceConfidenceExtractor
text → currentcurrentQwen — official blogT2mediumllm

Languageslanguages2

Claim history for Languages
ValueValid from → toStatusSourceConfidenceExtractor
92 programming languages → supersededQwen — official blogT2mediumllm
Chinese, English → currentcurrentQwen — official blogT2mediumllm

Official pageofficial_url1

Claim history for Official page
ValueValid from → toStatusSourceConfidenceExtractor
https://qwen.ai → currentcurrentQwen — official blogT2mediumllm

Paperpaper_url1

Claim history for Paper
ValueValid from → toStatusSourceConfidenceExtractor
https://arxiv.org/abs/2409.12186 → currentcurrentQwen — official blogT2mediumllm

Audioaudio1

Claim history for Audio
ValueValid from → toStatusSourceConfidenceExtractor
No → currentcurrentQwen — official blogT2mediumllm

Fine-tuning availablefine_tuning_available1

Claim history for Fine-tuning available
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentQwen — official blogT2mediumllm

Reasoningreasoning1

Claim history for Reasoning
ValueValid from → toStatusSourceConfidenceExtractor
Yes → currentcurrentQwen — official blogT2mediumllm

Training data notestraining_data_notes1

Claim history for Training data notes
ValueValid from → toStatusSourceConfidenceExtractor
5.5 trillion tokens including source code, text-code grounding data, and synthetic data, plus additional data on mathematics and general abilities. → currentcurrentQwen — official blogT2mediumllm

Visionvision1

Claim history for Vision
ValueValid from → toStatusSourceConfidenceExtractor
No → currentcurrentQwen — official blogT2mediumllm

Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

Provenance

Attributed facts

22

Source tiers

T222

Freshest observation

13 h ago

Conflicts

None

Source documents 4

Source documents
SourceDocumentTypeTierLast observedSnapshots
Qwen — official blogqwenlm.github.io/blog/qwen2.5-coder newsT1· Official13 h ago1
Qwen — official blogqwenlm.github.io/blog/qwen2.5 newsT1· Official13 h ago1
Qwen — official blogqwenlm.github.io/blog/qwen2.5-coder-family newsT1· Official13 h ago1
Qwen — official blogqwenlm.github.io/blog/qwen3-coder newsT1· Official13 h ago1

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.

Data quality (61/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →