Skip to content
llmwise
Sign inStart free

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 freeSee pricing

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

Short answer

In our test runs on September 27, 2026, DeepSeek V4 Pro passed 4 of 5 of our coding prompts (1 of the 2 hard ones): 14th of 15 models. Each reply counted as 1 message on Pro, where it has 250 a month, and cost $0.0132 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
PromptResultCostTime
Turn a title into a URL slugEverydayPassed: All 7 tests passed.$0.00132.1 s
Parse a duration like “1h 30m”EverydayPassed: All 10 tests passed.$0.003020.1 s
Merge overlapping intervalsEverydayPassed: All 7 tests passed.$0.00056.6 s
Evaluate an arithmetic expression, no evalHardPassed: All 15 tests passed.$0.029140.3 s
Parse CSV with quoted fieldsHardFailed: No answer within Pro's reply limit of 8,000 tokens: the model spent them all reasoning.$0.032369.1 s

Every coding reply from DeepSeek V4 Pro

  1. 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 title
    .normalize('NFD')
    .replace(/[\u0300-\u036f]/g, '')
    .toLowerCase()
    .replace(/[^a-z0-9]+/g, '-')
    .replace(/^-+|-+$/g, '');
}
```

    494 tokens in, 339 out (263 of them reasoning) · 2.1 s · $0.0013 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer ·

  2. 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) {
  const match = /^\s*(?:(\d+)h)?\s*(?:(\d+)m)?\s*(?:(\d+)s)?\s*$/.exec(text);
  if (!match || (!match[1] && !match[2] && !match[3])) return null;
  return (
    Number(match[1] || 0) * 3600 +
    Number(match[2] || 0) * 60 +
    Number(match[3] || 0)
  );
}
```

    477 tokens in, 1,544 out (1,420 of them reasoning) · 20.1 s · $0.0030 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Alibaba ·

  3. Merge overlapping intervals

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

    Passed: All 7 tests passed.

    ```javascript
function mergeIntervals(intervals) {
  if (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].slice()];

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

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

  return merged;
}
```

    461 tokens in, 457 out (283 of them reasoning) · 6.6 s · $0.0005 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via StreamLake ·

  4. 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) {
  function tokenize(input) {
    const tokens = [];
    let i = 0;

    while (i < input.length) {
      const ch = input[i];

      if (/\s/.test(ch)) {
        i++;
        continue;
      }

      if (ch >= '0' && ch <= '9') {
        const start = i;
        while (i < input.length && input[i] >= '0' && input[i] <= '9') {
          i++;
        }
        if (i < input.length && input[i] === '.') {
          i++;
          if (i < input.length && input[i] >= '0' && input[i] <= '9') {
            while (i < input.length && input[i] >= '0' && input[i] <= '9') {
              i++;
            }
          } else {
            throw new Error('Invalid number');
          }
        }
        tokens.push({ type: 'number', value: parseFloat(input.slice(start, i)) });
        continue;
      }

      if (ch === '+' || ch === '-' || ch === '*' || ch === '/' || ch === '^') {
        tokens.push({ type: 'op', value: ch });
        i++;
        continue;
      }

      if (ch === '(') {
        tokens.push({ type: 'lparen' });
        i++;
        continue;
      }

      if (ch === ')') {
        tokens.push({ type: 'rparen' });
        i++;
        continue;
      }

      throw new Error(`Invalid character "${ch}"`);
    }

    return tokens;
  }

  const tokens = tokenize(expression);
  let pos = 0;

  function peek() {
    return tokens[pos];
  }

  function next() {
    return tokens[pos++];
  }

  function isOp(value) {
    const token = peek();
    return token !== undefined && token.type === 'op' && token.value === value;
  }

  function parsePrimary() {
    const token = peek();

    if (token === undefined) {
      throw new Error('Unexpected end of expression');
    }

    if (token.type === 'number') {
      next();
      return token.value;
    }

    if (token.type === 'lparen') {
      next();
      const value = parseExpr();
      const closing = peek();

      if (closing === undefined || closing.type !== 'rparen') {
        throw new Error('Expected closing parenthesis');
      }

      next();
      return value;
    }

    throw new Error(`Unexpected token "${token.value}"`);
  }

  function parseExponent() {
    let base = parsePrimary();

    if (isOp('^')) {
      next();
      const exponent = parseUnary();
      base = Math.pow(base, exponent);
    }

    return base;
  }

  function parseUnary() {
    if (isOp('-')) {
      next();
      const operand = parseUnary();
      return -operand;
    }

    return parseExponent();
  }

  function parseTerm() {
    let left = parseUnary();

    while (true) {
      if (isOp('*')) {
        next();
        const right = parseUnary();
        left *= right;
      } else if (isOp('/')) {
        next();
        const right = parseUnary();
        left /= right;
      } else {
        break;
      }
    }

    return left;
  }

  function parseExpr() {
    let left = parseTerm();

    while (true) {
      if (isOp('+')) {
        next();
        const right = parseTerm();
        left += right;
      } else if (isOp('-')) {
        next();
        const right = parseTerm();
        left -= right;
      } else {
        break;
      }
    }

    return left;
  }

  const result = parseExpr();

  if (pos !== tokens.length) {
    throw new Error('Unexpected trailing tokens');
  }

  return result;
}
```

    563 tokens in, 8,285 out (7,436 of them reasoning) · 40.3 s · $0.0291 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via Wafer ·

  5. Parse CSV with quoted fields

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

    Failed: No answer within Pro's reply limit of 8,000 tokens: the model spent them all reasoning.

    (no reply)

    493 tokens in, 8,000 out (7,805 of them reasoning) · 69.1 s · $0.0323 · 1 message on Pro · answered by deepseek/deepseek-v4-pro-0813 via CoreWeave ·

How it compares on coding

DeepSeek V4 Pro beside the job's picks among every model. 13 models share the hard-problems pick, and DeepSeek V4 Pro isn't one of them.

DeepSeek V4 Pro beside other models on coding
ModelPassedOn ProCost per reply
GLM 5.3Our pick: hard problems (shared), best value5 of 5250 a month on Pro$0.0015
GLM 5.3 FlashOur pick: hard problems (shared), everyday5 of 560 a day on Pro$0.0003
DeepSeek V4 Pro4 of 5250 a month on Pro$0.0132
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: 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.

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.

DeepSeek V4 Pro, and coding, elsewhere

Questions

Is DeepSeek V4 Pro good for coding?

In our test runs it passed 4 of 5 coding prompts, 14th of the 15 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, and GLM, 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.

Start freeSee pricing