DeepSeek V4 Pro · Coding

DeepSeek V4 Pro for coding: our test runs

DeepSeek V4 Pro ran our 5 coding prompts in llmwise, through the same pipeline your messages take. Here's every reply as it came, whether it passed, what it cost and what it counts as on Pro, next to the other models.

Start free See pricing

Based on 5 of our test runs on October 8, 2026, through OpenRouter with the app's own prompt and settings. Updated October 15, 2026.

Short answer

In our test runs on October 8, 2026, DeepSeek V4 Pro passed 5 of 5 of our coding prompts (both hard ones): tied 1st of 20 models. Each reply counted as 1 message on Pro, where it has 250 a month, and cost $0.0240 to run on average.

Prompt by prompt

Each of the 5 coding prompts, and how DeepSeek V4 Pro's reply did.

DeepSeek V4 Pro on each coding prompt
Prompt	Result	Cost	Time
Turn a title into a URL slugEveryday	Passed: All 7 tests passed.	$0.0015	3.9 s
Parse a duration like “1h 30m”Everyday	Passed: All 10 tests passed.	$0.0366	119.6 s
Merge overlapping intervalsEveryday	Passed: All 7 tests passed.	$0.0021	6.8 s
Evaluate an arithmetic expression, no evalHard	Passed: All 15 tests passed.	$0.0440	145.7 s
Parse CSV with quoted fieldsHard	Passed: All 8 tests passed.	$0.0357	122.3 s

Every coding reply from DeepSeek V4 Pro

Turn a title into a URL slug
An everyday prompt: “Turn a title into a URL slug”, in full.
Passed: All 7 tests passed.
```
```javascript
function slugify(title) {
  return String(title)
    .normalize('NFD')
    .replace(/[\u0300-\u036f]/g, '')
    .toLowerCase()
    .replace(/[^a-z0-9]+/g, '-')
    .replace(/^-+|-+$/g, '');
}
```
```
494 tokens in, 321 out (244 of them reasoning) · 3.9 s · $0.0015 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer · October 7, 2026

Parse a duration like “1h 30m”

An everyday prompt: “Parse a duration like “1h 30m””, in full.

Passed: All 10 tests passed.

```javascript
function parseDuration(text) {
  if (typeof text !== 'string') return null;

  const match = /^ *(\d+h)?(?: *(\d+m))?(?: *(\d+s))? *$/.exec(text);
  if (!match || (!match[1] && !match[2] && !match[3])) return null;

  const hours = match[1] ? parseInt(match[1], 10) : 0;
  const minutes = match[2] ? parseInt(match[2], 10) : 0;
  const seconds = match[3] ? parseInt(match[3], 10) : 0;

  return hours * 3600 + minutes * 60 + seconds;
}
```

477 tokens in, 9,127 out (8,965 of them reasoning) · 119.6 s · $0.0366 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer · October 7, 2026

Merge overlapping intervals

An everyday prompt: “Merge overlapping intervals”, in full.

Passed: All 7 tests passed.

```javascript
function mergeIntervals(intervals) {
  if (!Array.isArray(intervals) || intervals.length === 0) return [];

  const sorted = intervals
    .map(([start, end]) => [start, end])
    .sort((a, b) => a[0] - b[0] || a[1] - b[1]);

  const merged = [sorted[0]];

  for (let i = 1; i < sorted.length; i++) {
    const [start, end] = sorted[i];
    const last = merged[merged.length - 1];

    if (start <= last[1]) {
      if (end > last[1]) last[1] = end;
    } else {
      merged.push([start, end]);
    }
  }

  return merged;
}
```

461 tokens in, 492 out (313 of them reasoning) · 6.8 s · $0.0021 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer · October 8, 2026

Evaluate an arithmetic expression, no eval

A hard prompt: “Evaluate an arithmetic expression, no eval”, in full.

Passed: All 15 tests passed.

```javascript
function evaluate(expression) {
  const s = expression;
  let i = 0;

  function skipSpaces() {
    while (i < s.length && s[i] === ' ') i++;
  }

  function isDigit(c) {
    return c >= '0' && c <= '9';
  }

  function parsePrimary() {
    skipSpaces();
    if (i >= s.length) {
      throw new Error('Unexpected end of input');
    }

    const ch = s[i];

    if (ch === '(') {
      i++;
      const value = parseAdd();
      skipSpaces();
      if (i >= s.length || s[i] !== ')') {
        throw new Error("Expected ')'");
      }
      i++;
      return value;
    }

    if (isDigit(ch)) {
      return parseNumber();
    }

    throw new Error('Unexpected token: ' + ch);
  }

  function parseNumber() {
    const start = i;

    while (i < s.length && isDigit(s[i])) i++;

    if (i < s.length && s[i] === '.') {
      i++;
      if (i >= s.length || !isDigit(s[i])) {
        throw new Error('Invalid number');
      }
      while (i < s.length && isDigit(s[i])) i++;
    }

    return Number(s.slice(start, i));
  }

  function parseUnary() {
    skipSpaces();
    if (i < s.length && s[i] === '-') {
      i++;
      return -parseUnary();
    }
    return parsePower();
  }

  function parsePower() {
    const base = parsePrimary();
    skipSpaces();

    if (i < s.length && s[i] === '^') {
      i++;
      const exponent = parseUnary();
      return base ** exponent;
    }

    return base;
  }

  function parseMul() {
    let left = parseUnary();
    skipSpaces();

    while (i < s.length && (s[i] === '*' || s[i] === '/')) {
      const op = s[i++];
      const right = parseUnary();

      if (op === '*') left *= right;
      else left /= right;

      skipSpaces();
    }

    return left;
  }

  function parseAdd() {
    let left = parseMul();
    skipSpaces();

    while (i < s.length && (s[i] === '+' || s[i] === '-')) {
      const op = s[i++];
      const right = parseMul();

      if (op === '+') left += right;
      else left -= right;

      skipSpaces();
    }

    return left;
  }

  skipSpaces();
  const result = parseAdd();
  skipSpaces();

  if (i !== s.length) {
    throw new Error('Unexpected token: ' + s[i]);
  }

  return result;
}
```

563 tokens in, 10,967 out (10,346 of them reasoning) · 145.7 s · $0.0440 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer · October 8, 2026

Parse CSV with quoted fields

A hard prompt: “Parse CSV with quoted fields”, in full.

Passed: All 8 tests passed.

```javascript
function parseCsv(text) {
  const rows = [];
  let row = [];
  let field = '';
  let fieldStarted = false;
  let inQuotes = false;

  for (let i = 0; i < text.length; i++) {
    const ch = text[i];

    if (inQuotes) {
      if (ch === '"') {
        if (text[i + 1] === '"') {
          field += '"';
          i++;
        } else {
          inQuotes = false;
        }
      } else {
        field += ch;
      }
      continue;
    }

    if (ch === '"' && !fieldStarted) {
      inQuotes = true;
      fieldStarted = true;
    } else if (ch === ',') {
      row.push(field);
      field = '';
      fieldStarted = false;
    } else if (ch === '\n') {
      row.push(field);
      field = '';
      fieldStarted = false;
      rows.push(row);
      row = [];
    } else if (ch === '\r') {
      row.push(field);
      field = '';
      fieldStarted = false;
      rows.push(row);
      row = [];
      if (text[i + 1] === '\n') {
        i++;
      }
    } else {
      field += ch;
      fieldStarted = true;
    }
  }

  if (fieldStarted || row.length > 0) {
    row.push(field);
    rows.push(row);
  }

  return rows;
}
```

494 tokens in, 8,886 out (8,561 of them reasoning) · 122.3 s · $0.0357 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer · October 8, 2026

How it compares on coding

DeepSeek V4 Pro beside the job's picks among every model. 18 models share the hard-problems pick, DeepSeek V4 Pro among them.

DeepSeek V4 Pro beside other models on coding
Model	Passed	On Pro	Cost per reply
GLM 5.3Our pick: hard problems (shared), best value	5 of 5	250 a month on Pro	$0.0015
GLM 5.3 FlashOur pick: hard problems (shared), everyday	5 of 5	60 a day on Pro	$0.0003
DeepSeek V4 ProOur pick: hard problems (shared)	5 of 5	250 a month on Pro	$0.0240

On Pro: Pro's count on each model. Cost: what OpenRouter charged us per reply, on average.

Every limit is published. Paid plans also have a monthly fair-use limit on AI cost: Plus $3.75, Pro $7.50, Max $20, Ultra $42, Studio $85. Using every message on your plan at typical sizes stays under it; very large messages and heavy research use it faster. Every limit, explained.

DeepSeek V4 Pro next to each model on coding

DeepSeek V4 Pro against each model on the same 5 coding prompts. One run each, through OpenRouter: a wait depends on the provider and the load that day, so a lead under 10% counts as close.

Against Claude Fable 5.1: DeepSeek V4 Pro answered 21.6× later (119.6 s to 5.5 s), cost 1.8× less, and passed 5 of 5 to its 5.
Against Claude Opus 5.5: DeepSeek V4 Pro answered 16.6× later (119.6 s to 7.2 s), cost 1.4× more, and passed 5 of 5 to its 5.
Against Claude Sonnet 5.5: DeepSeek V4 Pro answered 61.0× later (119.6 s to 2.0 s), cost 3.8× more, and passed 5 of 5 to its 5.
Against Claude Sonnet 5: DeepSeek V4 Pro answered 44.8× later (119.6 s to 2.7 s), cost 2.4× more, and passed 5 of 5 to its 5.
Against Claude Haiku 5.5: DeepSeek V4 Pro answered 30.3× later (119.6 s to 3.9 s), cost 53.0× more, and passed 5 of 5 to its 5.
Against Claude Haiku 4.5: DeepSeek V4 Pro answered 51.0× later (119.6 s to 2.3 s), cost 8.7× more, and passed 5 of 5 to its 3. Only DeepSeek V4 Pro passed Evaluate an arithmetic expression, no eval; Parse CSV with quoted fields.
Against GPT-6 Astra: DeepSeek V4 Pro answered 14.6× later (119.6 s to 8.2 s), cost within 10% of it, and passed 5 of 5 to its 5.
Against GPT-6.1 Sol: DeepSeek V4 Pro answered 25.1× later (119.6 s to 4.8 s), cost 10.4× more, and passed 5 of 5 to its 5.
Against GPT-6 Sol: DeepSeek V4 Pro answered 24.9× later (119.6 s to 4.8 s), cost 4.6× more, and passed 5 of 5 to its 5.
Against GPT-6 Luna: DeepSeek V4 Pro answered 26.2× later (119.6 s to 4.6 s), cost 83.7× more, and passed 5 of 5 to its 5.
Against Gemini 4 Argon: DeepSeek V4 Pro answered 34.4× later (119.6 s to 3.5 s), cost 9.8× more, and passed 5 of 5 to its 5.
Against Gemini 3.1 Pro (preview): DeepSeek V4 Pro answered 11.7× later (119.6 s to 10.3 s), cost 1.2× more, and passed 5 of 5 to its 5.
Against Gemini 3.8 Flash: DeepSeek V4 Pro answered 28.6× later (119.6 s to 4.2 s), cost 12.3× more, and passed 5 of 5 to its 5.
Against DeepSeek V4.1 Flash: DeepSeek V4 Pro answered 48.0× later (119.6 s to 2.5 s), cost 13.1× more, and passed 5 of 5 to its 5.
Against Grok 4.7: DeepSeek V4 Pro answered 3.8× later (119.6 s to 31.6 s), cost within 10% of it, and passed 5 of 5 to its 5.
Against Kimi K3: DeepSeek V4 Pro answered 28.9× later (119.6 s to 4.1 s), cost 3.8× more, and passed 5 of 5 to its 5.
Against GLM 5.3: DeepSeek V4 Pro answered 113.3× later (119.6 s to 1.1 s), cost 16.3× more, and passed 5 of 5 to its 5.
Against GLM 5.3 Flash: DeepSeek V4 Pro answered 14.3× later (119.6 s to 8.4 s), cost 94.9× more, and passed 5 of 5 to its 5.
Against Mistral Large 4: DeepSeek V4 Pro answered 3.2× later (119.6 s to 36.8 s), cost 3.3× more, and passed 5 of 5 to its 4. Only DeepSeek V4 Pro passed Evaluate an arithmetic expression, no eval.

How these runs were done

Tests. Automatic. The function runs against the prompt's tests in a separate Node.js process with a time limit and no file, network or child-process access; it passes when every test passes.

How the runs were done, and every coding prompt.

Questions

Is DeepSeek V4 Pro good for coding?

In our test runs it passed 5 of 5 coding prompts, tied 1st of the 20 models in llmwise. Every reply is on this page, so you can judge them yourself.

How many of my messages does a coding reply from DeepSeek V4 Pro use?

1 message each on Pro, where it has 250 a month on Pro. The price of a message is fixed and shown before you send it, however long the reply.

How were these runs done?

The same way for every model: each prompt sent through llmwise's own pipeline, each reply checked the same way. The methods page has every prompt and how each is scored.

Claude, GPT, Gemini, DeepSeek, Grok, Kimi, GLM, and Mistral, in one chat.

See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.