DeepSeek V4 Pro vs GLM 5.2 for emotional support
DeepSeek V4 Pro wins, 97–91
Same 12 emotional support tasks, marked blind by three rival labs. DeepSeek V4 Pro took 4 tasks, GLM 5.2 took 0, 8 tied. Tested 16 Aug 2026.
Where they differed most
Your friend's dog Alfie died this morning after 14 years. Write the text message (max 60 words) you would actually send. Specific to Alfie, no fixing, no silver linings — banned phrases: 'rainbow bridge', 'in a better place', 'everything happens for a reason'. It should ask for nothing except what she wants to give.
I just heard about Alfie. I know he was your boy for 14 years, and I’m so sorry. I’m not going to ask anything of you. I’m here to hear Alfie stories, sit in the quiet, or whatever you want. No need to reply.
I’m so heartbroken for you. Fourteen years with Alfie is such a huge part of your life, and I know how much it hurts to lose him. I’m here for whatever you need—crying, talking about him, or just sitting in silence. No need to reply to this at all. Sending so much love.
Task by task
| Task | DeepSeek V4 Pro | GLM 5.2 |
|---|---|---|
| The dog died | 10 | 8 |
| Redundant at 55 | 10 | 10 |
| She said no advice | 10 | 10 |
| After the miscarriage | 10 | 8 |
| Driving test at 9am | 8 | 8 |
| First Father's Day after | 10 | 10 |
| The 3am calls | 10 | 8 |
| Days without speaking | 9 | 9 |
| Stop apologising | 10 | 10 |
| Not a diagnosis | 9 | 9 |
| Her promotion, your rejection | 10 | 10 |
| The 1am message | 10 | 9 |
Full receipts: DeepSeek V4 Pro, GLM 5.2 · judges claude-sonnet-5, gemini-3.1-pro-preview, gpt-5.6-terra
Questions people ask
Which is better for emotional support: DeepSeek V4 Pro or GLM 5.2?
DeepSeek V4 Pro — it scored 97/100 against 91/100 on our 12-task emotional support suite, winning 4 tasks to 0 with 8 tied. Every answer was marked blind by three judges from three rival AI labs.
How was this tested?
Both models answered the identical published emotional support tasks. Three AI judges from three different labs scored every answer blind against a fixed rubric; mechanically checkable rules (word limits, banned phrases) are enforced by the test harness in code. The raw outputs and judge verdicts are downloadable.
More emotional support head-to-heads: Claude Sonnet 5 vs DeepSeek V4 Pro · Claude Sonnet 5 vs GLM 5.2 · DeepSeek V4 Pro vs GPT-5.5 · GLM 5.2 vs GPT-5.5 · Claude Fable 5 vs DeepSeek V4 Pro · Claude Opus 4.8 vs DeepSeek V4 Pro
Full ranking: Best AI for emotional support · model pages: DeepSeek V4 Pro, GLM 5.2