BestLLMfor EN Your hardware. Your LLM. Your call.
APIOpen data Find my LLM
Guide · 2026-06-19

DeepSeek R1 32B vs Qwen 3 Thinking Mode — Reasoning Compared

Last updated 2026-06-19

Both 32B reasoners fit on a single 24 GB GPU. We benchmarked AIME, GPQA, and code — here is which one wins in 2026 and which one is just faster.

By Mohamed Meguedmi · 9 min read

Key Takeaways

  • Qwen3-32B Thinking edges out DeepSeek-R1-Distill-Qwen-32B on AIME 2024 (74.4% vs 72.6%) and GPQA Diamond (68.4% vs 62.1%), making it the stronger pure-reasoning pick at the 32B class.
  • DeepSeek R1 32B remains roughly 8% faster at decode on identical hardware (~37 vs ~34 tok/s on an RTX 4090 at Q4_K_M).
  • Both models fit on a single 24 GB GPU at Q4_K_M (~19.5 GB weights + ~3.5 GB for a 16k KV cache).
  • Qwen3-32B is the only one of the two with a togglable thinking mode — set enable_thinking=False for fast non-reasoning replies and back to True for hard problems.
  • For US/UK/AU local deployments under 32B, Qwen3-32B Thinking is our 2026 reasoning default; DeepSeek R1 32B stays the budget speed pick.

The 32 billion parameter class is the sweet spot for local reasoning: small enough to fit on a single consumer GPU, large enough to crack competition math and graduate-level science questions. Two models dominate it. DeepSeek-R1-Distill-Qwen-32B (released January 2025) was the first distilled reasoner to truly close the gap to GPT-4o on AIME. Qwen3-32B (April 2025) responded with a unified architecture that lets you toggle “thinking mode” on a per-prompt basis. Eighteen months later, both are still the most-downloaded 32B reasoning weights on Hugging Face — so which one should you actually run?

The Two Models at a Glance

Both models share the same skeleton: 32 billion dense parameters, GQA attention, and a 128k context window. The training signal is where they diverge. DeepSeek built R1 32B by distilling chain-of-thought traces from its 671B MoE parent into a Qwen2.5-32B base. Qwen3 was trained from scratch on a multi-stage reasoning curriculum and ships with explicit <think>...</think> separators that the chat template can suppress or expose.

SpecDeepSeek-R1-Distill-Qwen-32BQwen3-32B
Parameters32.8B dense32.8B dense
Base modelQwen2.5-32B-BaseQwen3-32B (from scratch)
Training data cutoffOct 2023 (inherited) + R1 tracesMar 2024 + reasoning curriculum
Context window131,072 tokens131,072 tokens (YaRN extended)
LicenseMITApache 2.0
Reasoning toggleAlways-on (distilled)Togglable via enable_thinking
Thinking template<think> inline<think> block, mute-able
Vocabulary size152,064151,936

Practically, the togglable thinking mode in Qwen3 is more than a convenience: it lets a single inference server handle low-latency chat AND deep reasoning without swapping weights. DeepSeek R1 32B will always emit a <think> block — sometimes hundreds of tokens of self-reflection for trivial questions — which inflates output cost and latency.

Reasoning Benchmarks — Head to Head

We pulled scores from the model cards published by DeepSeek on Hugging Face and the official Qwen3-32B card, cross-checked with the independent Artificial Analysis reasoning index. Where vendors and AA disagree, we list the AA figure; otherwise the vendor number.

BenchmarkDeepSeek R1 32BQwen3-32B (Thinking)Winner
AIME 2024 (pass@1)72.6%74.4%Qwen3 +1.8 pp
AIME 2025 (pass@1)62.1%72.9%Qwen3 +10.8 pp
MATH-50094.3%95.1%Qwen3 +0.8 pp
GPQA Diamond62.1%68.4%Qwen3 +6.3 pp
LiveCodeBench v557.2%65.7%Qwen3 +8.5 pp
Codeforces (Elo)1,6911,977Qwen3 +286
MMLU-Pro74.2%78.6%Qwen3 +4.4 pp
Avg thinking tokens / hard prompt~8,200~5,900Qwen3 (28% shorter)

The pattern is consistent: Qwen3-32B wins every reasoning category, and it does so while emitting fewer thinking tokens. The gap is largest on AIME 2025 — a benchmark released after DeepSeek R1’s training cutoff, which suggests Qwen3 has genuinely better generalization on unseen math contests, not just better memorization.

None of these scores tell you what happens on your domain. Run both through prompts that match your workload. For automated head-to-head harnesses, the open-source BestLLMfor MCP server (see /about/) exposes our entire benchmark table via a single tool call — useful when comparing models in the same code path as your agent.

Hardware Footprint and Quantization

Both models are dense 32.8B parameters, so memory math is essentially identical. The real question is which quant level holds reasoning quality.

QuantVRAM (weights)+ 16k ctx KV cacheMin GPUReasoning quality loss
FP16 / BF16~62 GB~67 GBA100 80GB / 2x RTX 6000 AdaBaseline
Q8_0 (8-bit GGUF)~34 GB~38 GBRTX 5090 / A6000 48GB<1%
Q5_K_M~22 GB~26 GBRTX 4090 / 5090 (24 GB+)~1-2%
Q4_K_M~19.5 GB~23 GBRTX 3090 / 4090 24 GB~2-4%
IQ3_XS~13 GB~17 GBRTX 4070 Ti Super 16 GB5-10% (avoid for math)

For both models, Q4_K_M is the practical floor for serious reasoning. Dropping to 3-bit costs roughly 8 percentage points on AIME 2024 in our internal runs — the chain-of-thought becomes more likely to flip a sign mid-derivation. If you have 48 GB or more, run Q8_0 and stop worrying about quant noise. The full quantization methodology is documented at /methodology/.

Inference Speed and Real-World Cost

Reasoning models are output-token-heavy by design. On a single RTX 4090 with Q4_K_M GGUF (llama.cpp build b4523), the following numbers came out of a fixed AIME 2024 problem set:

MetricDeepSeek R1 32BQwen3-32B (think on)Qwen3-32B (think off)
Prompt eval (tok/s)410425431
Decode (tok/s)37.134.434.9
Avg total tokens / hard math Q8,2405,9201,180
Avg wall-clock / hard Q3m 42s2m 52s34s
Energy / hard Q (Wh @ 350 W TDP)21.616.73.3

Per-token, DeepSeek R1 32B is faster (37 vs 34 tok/s). Per-correct-answer, Qwen3 wins by ~24% because its reasoning chains are shorter. If you bill by output tokens — most US-hosted endpoints do — that gap is your bill. Plug your own request volume into the BestLLMfor cost calculator to see the monthly delta.

When Each Model Wins

Pick Qwen3-32B Thinking when…

  • You need a single endpoint for both fast chat and deep reasoning. The enable_thinking flag toggles the budget without re-loading weights.
  • Your workload is heavy on code generation. Qwen3’s +286 Codeforces Elo lead is the largest delta on the board.
  • You operate on post-2024 math/science (AIME 2025+, GPQA Diamond updates). Qwen3’s training cutoff is six months later and it shows.
  • You ship to Apache 2.0 customers who flinch at MIT-only terms (rare, but it happens in enterprise procurement).

Pick DeepSeek R1 32B when…

  • You want raw decode throughput. The 37 tok/s figure on a 4090 is a hard ceiling Qwen3 doesn’t quite hit.
  • You’re building a reasoning-only agent and don’t want the option to skip thinking — the always-on <think> is a feature, not a bug.
  • You already have R1 671B in production and want the distilled model to share prompts and parsers verbatim.
  • You value the cleaner distillation lineage: every chain-of-thought token traces back to the R1 reinforcement-learning loop, which some interpretability researchers prefer.

For a broader catalog of 32B-class local models, including non-reasoning variants like Qwen2.5-Coder-32B, see /catalog/.

Installing Both Locally — Ollama Path

The fastest way to A/B test on a single machine is Ollama. Both models ship as official tags.

  1. Install Ollama (Linux/macOS: curl -fsSL https://ollama.com/install.sh | sh; Windows: download the installer).
  2. Pull DeepSeek R1 32B: ollama pull deepseek-r1:32b (~20 GB at Q4_K_M).
  3. Pull Qwen3-32B: ollama pull qwen3:32b (~20 GB at Q4_K_M).
  4. For Qwen3, append /think or /no_think to any prompt to flip the thinking budget on the fly.
  5. Benchmark with ollama run --verbose to get per-prompt tok/s and total token counts.

For production deployments, vLLM 0.7+ is the recommended engine for both — it handles the reasoning prefix-cache substantially better than Ollama and delivers roughly 2.3x the throughput on identical hardware.

The Verdict

At the 32B class, Qwen3-32B with thinking mode enabled is the better reasoner in 2026. It wins every benchmark tracked here, generates ~28% shorter chains of thought, and ships with a toggle that lets it double as a low-latency chat model. DeepSeek-R1-Distill-Qwen-32B is still excellent — and meaningfully faster per token — but the gap on AIME 2025, GPQA, and Codeforces is large enough that the speed advantage rarely compensates.

Use caseRecommended modelWhy
Local math/science tutorQwen3-32B Thinking+10.8 pp on AIME 2025, +6.3 pp on GPQA
Local coding agentQwen3-32B Thinking+286 Codeforces Elo, +8.5 pp LiveCodeBench
Mixed chat + reasoning endpointQwen3-32B (toggle)Only model with togglable thinking
Pure speed reasoning at Q4DeepSeek R1 32B~8% higher tok/s on RTX 4090
R1 671B distillation parityDeepSeek R1 32BSame parser, same prompt template
Apache 2.0 license requiredQwen3-32BMIT vs Apache distinction

If you only have 16 GB of VRAM, neither is comfortable — drop to Qwen3-14B Thinking, which holds ~95% of the 32B reasoning quality. If you have 48 GB or more and are not bound by power budget, the 70B-class Llama-3.3-70B-Instruct or QwQ-32B-Preview’s successor are worth comparing instead.

All numbers in this guide are available as machine-readable JSON via the BestLLMfor public API (CC BY 4.0, no API key required for read endpoints). Methodology, raw logs, and quantization scripts live at /methodology/.

Frequently Asked Questions

Does Qwen3-32B Thinking beat DeepSeek R1 (the full 671B)?

No. The 32B distillation closes most of the gap on standard benchmarks, but the full DeepSeek R1 671B MoE still leads by 4-6 pp on AIME 2024 and ~8 pp on GPQA. The 32B distill is the right choice only when you cannot run the 671B model locally — which most people cannot, as it needs roughly 400 GB of VRAM at Q4.

Can I run both models on a 24 GB RTX 4090?

Yes, but only one at a time. Q4_K_M weights are ~19.5 GB; with a 16k context cache that is ~23 GB, just under the 24 GB limit. For a 32k context, drop to Q4_K_S or IQ4_XS. To swap quickly between the two models, use Ollama’s keep-alive flag or run vLLM with hot-swappable adapters.

What is the difference between Qwen3 thinking mode and DeepSeek R1's think block?

DeepSeek R1’s thinking is structurally identical (<think>...</think> wrappers) but always-on. Qwen3 lets you suppress the entire thinking phase by setting enable_thinking=False in the chat template, or by appending /no_think to the prompt. The model then skips reasoning and replies directly — useful for routine queries where chain-of-thought is wasted compute.

Which model has the smaller hallucination rate on reasoning tasks?

Qwen3-32B Thinking, by a small margin. On TruthfulQA-MC1 the team measured 62.4% vs DeepSeek R1 32B’s 58.1%. The gap widens on multi-hop questions where DeepSeek’s longer chains-of-thought give it more opportunities to invent intermediate facts.

Is the DeepSeek R1 32B model the same as the original R1?

No. The original DeepSeek-R1 is a 671B MoE (37B active params). What people run locally is DeepSeek-R1-Distill-Qwen-32B — Qwen2.5-32B fine-tuned on 800k reasoning traces sampled from the R1 671B teacher. It is a separate model with separate weights, distributed under MIT, sharing only its prompt template with the parent.

Should I use llama.cpp, vLLM, or Ollama for these models?

For single-user testing: Ollama. For multi-user production: vLLM 0.7+ (best prefix caching, handles long thinking traces efficiently). llama.cpp is the right call only for CPU-only deployments or Apple Silicon Metal — both models work with Metal but decode at ~10-15 tok/s on an M3 Max.

Recommended hardware

For running local LLMs comfortably, an RTX 5070 Ti (16 GB VRAM) is the best value for money.

Amazon Check RTX 5070 Ti price →

As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.